The fact that you end up losing all of your money if you're wrong, and end up earning a bunch of money if you're right, and that we can look at the histories of traders and prediction markets, kind of makes me less concerned about this
I think it would be hard to make a sock puppet into a mod, since a user generally needs to have a well-established history in the community to get trusted with a mod position, and an account that was created right when the previous admin decided to step down wouldn't have much history.
(And if you've been posting frequently for years on an account to create the appearance of a well-established user, that's less a sockpuppet and more a main account.)
I agree that if, eg, Trump found a way to hack the election machinery to ensure he'd win, he would win. And if at the same time he bets on himself winning in a prediction market, he could make money off of it. That seems like generic insider trading. Is that your point, or are you trying to say something more complicated?
as you can see from the OP, attempted manipulation can make a minor market the most popular market on the site due to all the people taking the manipulator's money. There's no market small enough for continuous manipulation to be cheap.
What if Trump bet $1B that he would lose the election, then resigned?
What if the person choosing the next moderator was the one who poured $3500 into the market predicting it would be a woman, then chose a woman?
What if a scientist is very close to discovering a cure for cancer and their efforts are highly publicized, and they notice the markets predict they will publish a cure in 6 months, so they spend all their money on 2-year predictions and then chill in Tahiti for a year and a half while people keep dying of cancer?
Etc.
In a sense, yes, this is the same as a CEO shorting their own company and then intentionally crashing it.
But I *think* it's already illegal for a CEO to do that? Or at least they would never work again? Whereas I don't see any penalty for doing that in prediction markets as they currently exist.
I think you're generally correct that, for whatever worry we could come up with about manipulating prediction markets and creating bad outcomes, there's some analogy you could draw to the same type of behavior in a stock market.
But I think stock markets are incredibly highly regulated in all sorts of ways in order to prevent those examples, *and also* stock market manipulation *does* happen and *does* lead to all kinds of really bad outcomes, fairly often.
It’s a good point. Presumably the SEC (or an equivalent entity) would eventually get involved and apply stock market-like insider-trading laws to these markets if they got big enough?
Ie you can only trade on public information, cannot directly manipulate the markets, etc. It seems that the same “fairness” intuitions we apply to stocks would apply here.
Which would be a shame since insider trading helps price discovery. In the Trump example above, the $1B entering the market is the smart/correct money...
I'm not sure this is the case. In the worst case we're talking about people changing their behavior after bidding on an unlikely outcome they can affect, not just insiders betting on outcomes that they have private information on.
For example I think the Trump imaginary scenario was sketching the idea that Trump was not planning to resign, but decided that a 10x yield (or whatever) on $1B was enough to convince him to do so.
I don't think the SEC would actually need to get involved. One high profile case of insider betting here would make a lot of potential bettors really paranoid about how the markets are working and who is betting what. (Who in the "Did one guy make those bets? I should look into that." sense, not the literal identity sense.) My guess is that such outcomes would become pretty self limiting as users dig into what's going on during the event, and the problems would be discovered and fixed before the regulators recognized a problem after the event and investigated.
If the insider was betting on what he planned to do, it actually would give good information, too. If someone wants to spend $10,000,000 to tell me ahead of time that they are going to tank a company, ok?
On the other hand, if they are trying to manipulate the information output because they want the company to tank based on misperceptions of its viability, that might be a problem. Maybe not, though, as other people who know about the company's viability would bet against them hard, which improves the information, and perception only goes so far towards affecting reality.
Edit: forgot the Trump thing. You might see some regulation around people starting the markets not being able to bet on the outcomes, or something like that. Possibly that would be imposed by the market makers themselves. I am given to understand that it is highly frowned upon for athletes to bet against themselves in contests, for example, though I don't have any idea how true or enforced that is.
Insider trading laws aren't about fairness, they're about theft. If you trade on misappropriated information (e.g., things you learn in your capacity as a CEO *belong* to the shareholders, and trading on it for your own benefit deprives them of value), that's a crime; if you learn something no one else knows through legitimate means, it's arguably unfair to other traders who don't know it but it's not a crime.
For two examples mentioned in this thread, if the person responsible for choosing the next moderator for the Discord was the one manipulating the market, it would make a ton of sense, and a make him a ton of 'cash' (though you've said elsewhere he claims it isn't him). For another example, in the question about whether the identity of the person who leaked the Roe v Wade docs will be known before 2023, if the 'No' side has a significantly large enough pot, the leaker could buy a ton of 'Yes' and make a killing by outing themselves.
For a bit more morbid example, I don't remember who said it first, but basically any prediction market is also a de facto assasination market. Buy enough 'No' positions on a 'will this person still be in power' contract, and you can make a killing, making a killing.
That's part of the point of a prediction market. If someone is trying to covertly manipulate the actual event, then someone who knows about it is likely to reveal this manipulation by buying up the market.
Looking at the first graph of the PredictIt market, 4% for the decision in Dobbs still seems awfully high. That's not based on any actual knowledge of the case or the court. It's based on the fact that "MI" stands for Michigan, not Mississippi.
2. I just checked the PredictIt page and it says "MS". I think the caption is courtesy of Scott and he flubbed it because he was in a rush.
3. PredictIt markets have official rules sections more detailed than the market title, so even if the title said MI the rules section would probably say the right state.
Not to say that PredictIt markets never end up resolving in a way contrary to the expectations of many market participants! I can think of a solid handful, all a lot of fun.
Missed a main point: $14000 can turn a completely tertiary question into the biggest market. Or in other words, these prediction markets are eminently manipulable for minimal cost. This is a recipe for f***ery.
I would think this generally allows someone to "make a mountain out of an anthill". The manipulator would probably lose money on the market, but can gain something (awareness/money/distraction) outside of the market. E.g. highlight a vague, but possible inoffensive point of a some law and make it into a national issue worth debating about during the election cycle. It's already possible, of course, with traditional media, astro-turfing on the net, etc., but prediction markets probably would offer yet another channel of attack on the public discourse. The market probably should correct such attacks quickly enough, though, unless they're backed by a cult-like movement which can sustainably bet along with the instigator, however irrationally (e.g. a TV preacher congregation willing to throw money wherever their leader points to).
But such attacks could only work reliably on small-scale markets, I think. Otherwise it would take too much money to make even a blip against really hot predictions.
And be lost in the general din of advertisement. Of course, the effectiveness of this tactic would depend a lot on popularity of predictions markets and the audience that will participate and track them.
Manifold is a play money market where nobody is expected to spend real cash, which is why someone actually spending that much made everything crazy. The point I draw is "even if one person has much more money than any other individual and spends it on manipulation, everyone else will gang up to de-manipulate it"
> Manifold is a play money market where nobody is expected to spend real cash
That's obviously the correct response here, but I've never quite understood why you don't think it undermines this whole enterprise. You seem very interested in the play money markets, at least partly as a source of information about how large-scale real-money markets could work. But surely the differences are too large to draw many serious conclusions about one from observing the other? If you've written about this in detail somewhere I'd be interested to read it.
I mostly do think it undermines the enterprise. I don't think it makes it 100% useless, but it makes it not great for anything with real stakes. I do think this current lesson still holds because it's a case where the play-money-ness ought to work against rather than for it, but it still succeeded.
I think the real issue is that right now, these are mostly a playground for making money or messing around.
The more these are used as real indicators of important events, the less useful they will become, because they will be more prone to both manipulation for the purpose of external manipulation, as well as because there will be more less informed people involved.
This is an issue with the real stock market - market manipulation can greatly manipulate the value of certain targeted stocks (like Gamestop or Tesla) but it's more difficult to manipulate the value of all stocks in general. However, if whatever the stock in question is happens to be important at that time, then the manipulated stock might be a big deal.
I envision a scenario of government fueled propaganda working in concert with prediction market manipulation (assuming the markets become relevant in the future).
So Putin (or whoever) says something about his plans, and has his propaganda machine back it up, and simultaneously has a hidden organization bid up the price as if what Putin is saying is correct. There's enough noise available to the public that it's harder to make a prediction, and the market itself seems to be backing up a particular conclusion, making it less likely people bet against it.
That assumes that the information isn't actually available to prove the case either way, but with sophisticated propaganda, that's often the case. In the specific instance of Trump bidding up the "Trump Wins" category with $100 million - right now people would be fools to bet $100 million against Trump winning in 2024. Is he definitely going to win? No, but there's a strong enough chance that it would be extremely dumb to bet that much money against it. Maybe the true percent should be 40%, but because of the size of the bets in favor, the market says 60%. So Trump's $100 million keeps the market floating, enhancing his own chances of actually winning (people like winners).
What if the manipulator were a bit more subtle? Instead of dumping thousands of dollars at once and shifting the market straight to 100%, you'd want to shift it slowly to avoid attracting attention to the market.
I think this is a highly atypical market anyway since it's too prone to insider trading. Without knowing anything about how the moderator is being selected, I would assume that anyone placing significant bets far away from the consensus is probably party to some inside information.
But doesn't a sudden market movement like that make people who *don't* know it should be at 10% take notice? Like yeah, the ones who have familiarity with and grasp of the subject matter would have noticed either way. But surely there are a whole host of people who'd simply have noticed that this random market was suddenly the biggest in the history of the site, and had jumped suddenly from one extreme to the other, and realised what had happened (or at least deemed it worthy of inquiry)?
Whereas if you do it more subtly and slowly, none of those people need be alerted.
That's what seems really weird to me. Insider trading is generally a *good* feature on prediction markets, because it gets you a *lot* more accuracy than just trading on general knowledge. In fact, getting insiders to reveal their private information is sometimes said to be one of the mean benefits of prediction markets.
If the market reacts this strongly to block someone spending a lot of money on it, wouldn't it also react this strongly to some insider trying to cash in on their knowledge? And wouldn't that defeat the purpose of the market?
If the insider trader has good intel and the market "fights" them on it, then it's the insider who gets to make a bunch of money when the question finally resolves.
The same could have been said for cryptocurrency (play money where nobody expected to spend real cash), so not so clear to me that your point is as strong as you evidently believe it is.
How are they analogous? From the beginning, cryptocurrencies were intended to have real-world value. I don't think anyone expects their prediction-market play money to ever be exchangeable for goods or services or dollars.
Cryptocurrencies were intended to play a real world role - that of currency.
In actual reality, cryptocurrencies were used initially primarily to evade capital controls plus facilitate cyber crime.
Prediction markets are intended to fulfill a real world role: prediction.
In reality, how will this be skewed by the unscrupulous and/or conniving?
This will be any place where prediction confers advantage.
Election expected outcome is a known influencer of turnout.
Conferring of credibility onto otherwise utter nonsense endeavors is another.
I'm am 100% certain there is an infinite portfolio of these since we already know the widespread impact of past prediction purveyors such as astrologists and palm readers.
I never participated in the "what do you most disagree with Scott about?" thread, and I agree with your points regarding why prediction markets seem useful. But, I think if there is something I do often disagree about, it's the usefulness of measures for turning prediction markets into *drivers* of policy. I think there are some cases where this is probably a good idea, but I think that in a lot of cases, attempting to make the step from using them as an analytical tool to using them as a policymaking tool runs headlong into Goodhart's Law.
Well said, that’s a situation where there would be clear incentives to manipulate the market in advantageous ways - like cutting out the middleman of having to bribe tons of congressmen with donations. I don’t see a way they can make policy that avoids this unless it’s on very narrow interests without finanical stake - but then what’s the point of such a specific tool?
I don’t know much about this and would be interested to read a steelman by Scott or another
So long as the mispricing is more than 5%+10%/year a lot of people will gang up to un-manipulate it. This happens regardless of the motivations of the manipulator, so I don't think it makes any difference whether the manipulator is motivated by the fact that the prices will be used for policymaking.
if they have more money than everybody else combined and really really want to throw it away manipulating a market, ok, but manipulation on that scale would be pretty obvious and people wouldn't trust the probabilities.
Will it really be so easy to identify manipulation? For any market where it's even plausible that someone has important non-public information, how can even the smart money tell the difference between 'person with inside knowledge places large, accurate bet' and 'manipulator places large, inaccurate bet'?
Unusually high volume, and common knowledge among the high ranking forecasters that it seems insanely mispriced. It wouldn't go unnoticed, but it could be hard to distinguish insider trading vs inefficiency vs manipulation. People would take the numbers with a grain of salt even if they can't determine which.
This reminds me of the betting market for who would get the iron throne. Enough insiders bet on Bran early to make him a -1000 favorite before the show aired, which looked crazy to anyone without the insider knowledge.
If you're asking a some forecasting experts whether the market should to be trusted or is mispriced, why bother with the market at all? You could just ask the experts what price they think it ought to be.
In this case the experts are determined objectively by their track records on the prediction market, instead of whichever clown the fourth generation of Sulzbergers wants to give an NYT column to.
I mostly agree with this - I'm hopeful that one day prediction markets could be involved in policy, but this would be after we understand them a lot better and have much more experience with them.
What do they get from taking it over? Unlike a pundit, you can't just choose to make the prediction market say anything you want (unless you make it stop being a functional prediction market). If it says a number that's wrong, people will make money by betting against it.
That's not so much "taking it over" as "destroying it". It would be like seizing control of a newspaper just so you could shut it down.
Certainly that's a thing that could happen, but it's extremely different from corrupting the newspaper to print favorable stories. It's not the same problem that Scott was talking about in that sentence you quoted.
(You can't _pretend_ that the prediction market is still functional, the way you could pretend that a newspaper is still impartial, because regular people can place bets in a prediction market and then make money if they're right. If you just made your web site claim the consensus was 90% when it was actually 10%, you'd get a ton of people betting on 10%. Then when the thing doesn't happen, you've either got to pay off all those bettors--in which case this wasn't too different from just placing a large bet on an honest market--or else refuse to pay them, in which case all the bettors will realize your so-called prediction market is a fraud.)
Imagine the US government creating thousands of sophisticated sock accounts with virtually unlimited funds. Then they place normal looking bets in a consistent direction, thoroughly pushing markets in a particular direction. At the same time, they release particularized subsets of real information at their disposal, creating an official source of data that backs up what they are betting on.
Is there any reason that they couldn't do this for things that are false as much as for things that are true? Sure, the US government might not want to blow millions of dollars on every question that comes up, but do you think a political party wouldn't be willing to spend huge amounts of money when they are in charge and the results of prediction markets have substantial affects? (Imagine a prediction market for abortion, with the outcome affecting a national law for or against). Bonus, if you can actually swing the outcome the way you want, you get your opponent's money as well as your own back!
Is this true? Has the US government taken over the NYSE or NASDAQ, in the sense that if a business they support's stock goes down, they can force NYSE/NASDAQ to make the stock go back up again?
I am a layman on the specific subject but that does not seem like a fair comparison; I would think that lack of explicit stock market manipulation is not because there are no avenues or desire for the US government to do so (simple imaginary path: well-targeted tightening or loosening of regulations) but rather because there's recognition that short-term stock price manipulation probably doesn't have all that much bearing on the long-term performance of whatever business the company actually wants to engage in.
There may be some analogous reason to not attempt manipulation of prediction markets but I don't see any at the moment; if they become sufficiently important, the current prediction itself exerts some non-total but non-negligible effect on opinions (whether of predictors or not), and it would seem useful to put a thumb on the scale in some cases. I don't think you can make a safe assumption that practical prediction markets will be immune to internal corruption and infiltration and the like, which will open up means of manipulation that are not possible within the regular rules of prediction markets.
Hmm, rereading that I recognize that there's a similar argument for putting a thumb on the scale for stock prices, but I would again attribute it to an understanding that it's not worth the squeeze, rather than not having any utility, and I would maintain that the cost and difficulty of manipulating a prediction market are likely to be lower than manipulating a major stock exchange.
What about all the "too big to fail" banks and the money dumped into supporting them? That may not be direct manipulation but neither is it letting the market sort itself out.
What's the analogy here? The government might pass a law causing the prediction to be true/false? But that's just how things normally work, the prediction market isn't involved at all
I think the analogy is that the government officials derived some personal utility by not letting the banks fail, and so used their official powers to make that happen. The government officials didn't force NASDAQ or DOW etc. to just write the stock prices higher, but they did force us all to pay a crap ton of money to make the stock prices higher via bail outs.
In terms of prediction markets, if the outcome of the market "Is policy X a good idea?" goes down, are officials who want X going to have an incentive to either spend money to make the market go towards YES, or perhaps more subtly, promise to change or actually change other rules such that X seems like a good idea, even if the rules changes are net negative? It might well devolve into a "I am going to break things until people agree that what I want is a good idea" situation. Admittedly, a lot of policy is like that, a series of kludges and band-aids on a system instead of ripping out or fixing the broken bits, but then that's why I can easily imagine a government manipulating a prediction market that way.
At the extreme, if prediction markets become important enough to drive policy, the government will start a "Board of Analytic Predictive Services" to control the spread of prediction market manipulation. Markets that the government deems manipulated, by virtue of making obviously false predictions in the eyes of the board members, will get BAPed until they display the right prediction.
Or more extreme, the government will just make their own market, make all others illegal, and then have a pricing algorithm such that the predictions are remarkably close to suggesting what the government officials wanted to do anyway.
Why go through a middleman in this way? Obama engineered the saving of GM.
This doesn't appear to be a good analogy, unless prediction markets could be circumvented to achieve the same policy goals (which is obviously the case right now in 2022, but we're talking about a future in which prediction markets serve some major purpose in society) and the government chose to go through prediction markets anyway.
Why don't these markets post odds in the traditional American odds format? Seems like either a silly oversight, or maybe it's just needless exclusionary. If these sites are interested in getting more traffic and action, then a first step would be post odds in a format that the typical gambler can understand.; i.e., if something has a 60% of happening then it's -150, or if it has a 20% chance of happening it's +400.
I think a tiny fraction of Americans understand American moneyline odds. Every newspaper article called Rich Strike an 80-1 long shot, not a +8000 one. (And that's not even getting into the fact that overseas people probably wouldn't be familiar with American odds either.) Most Americans understand what 60% means (or at least I'd like to think so).
Thanks. I'm British and grew up with the 'fractional odds' system, without even realising it had a name. It never occurred to me that other countries would use other systems.
Given all the Americans that are replying here saying they've never heard of the Money Line odds system, perhaps it shouldn't be referred to as "the American odds" system. (A bit like how 10^12 should probably no longer ever be referred to as a "British billion".)
Am American, this is the first time I've ever encountered this odds format. From your comment alone I can make no sense of it.
I've also never gambled, except for casually (as in "Bet you $10 you can't make that basketball shot" with my cousin sort of thing). What percentage of Americans gamble enough to be familiar with this kind of odds format?
I don't think they intend for their primary market to be regular non-prediction-market gamblers, and I think you might be suffering from some social myopia on how widely understood the traditional odds format is. I would wager (ha!) that among the general population, people who both understand it and prefer it to a % chance are a small minority; I would even make a weak wager that with the recent expansions in casual "nontraditional" gambling that people who understand and prefer this format are a minority even among people who gamble.
Probably because the designers and their expected early adopters weren't/aren't sports gamblers. The current format is transparent both to people who are familiar with basic statistics (contract price trivially translates to percentage probability) and people who are familiar with futures and options (you're spending the posted contract price to buy a thing that will have different values depending on what winds up happening. I had no idea how to read American sports betting odds until I looked it up just now, and I suspect the same is true of most other non-gamblers.
The current format, at least to me, is indecipherable. Taking the Ethereum market, for example-- I have no idea what 'no $.97' means, in terms of the risk/reward ratio. I'm not saying that the current layout isn't intuitive, only that at least to me it isn't. I think I can assume that the 'no' is the chalk, but that's as far as I can get.
Also- I'm sure you're right re: the intended audience, but if one of the goals is to create action it seems to me like it would make sense to offer odds in a few different formats that gamblers can understand. There are thousands of people out there who would happily take a $2000 flyer on a random prop, just for the hell of it, and the chances of drawing those people in would (I think) improve considerably if the odds were posted in a more conventional format. Most sportsbooks post odds in both American odds (for Americans) and decimal odds (for Europeans) for that reason, and doing something similar here would probably increase the liquidity of these props.
I have no familiarity with regular gambling nor prediction markets, but looking at the Polymarket format, I noted that it presents "No $0.97" in opposition to "Yes $0.03", adding up to $1, and I intuitively understood it to mean that the cost of a single unit of a wager (I don't know what the correct term for this is) on "no" is $0.97 and for "yes" is $0.03, and either would pay out $1 if it won.
Since you again casually used a term that only a tiny fraction of people would understand ("chalk"), I would guess that you are part of a demographic likely to interpret these presented formats in a very different way than the average layman, and I do think the prediction markets have aimed their odds presentations more at the average layman than at experienced gamblers.
Ha! That's fascinating. Thanks for sharing. So when you saw the layout it was clear that you were laying 97 cents to win a dollar on 'no', or laying 3 cents to win a dollar on 'yes'? I've looked at that graph for 15 minutes and that never occurred to me, though now of course that makes perfect sense.
To be fair it's not like I've never seen this format before in passing, but today is probably the first time I actually tried to interpret it. I don't know that it's *the* most intuitive format but I think it's at least a bit more newbie-friendly than the regular American odds format, for which I'd be in the same situation as you - I'd take forever to figure it out without looking it up.
I can totally understand that, now that you've brought it up. And there does seem to be something about betting odds specifically that makes them exceptionally difficult to interpret in unfamiliar formats, even when the format in question has been explained to us (I don't know if this is true for other numerical representations, though it seems likely that it is). For example, my wife-- who is not a gambler-- finds European decimal odds intuitively easy to understand, while I still struggle with them after twenty years ( I have to convert them into a percentage, and then convert them into American odds, before I have a clear picture of what I'm being asked to bet on).
Bad optics (association with gambling is bad) and also hell no that is horrible scheme for 97% of the world. Not trying to be mean but this idea is so bad I can't tell if it's a joke or not
I don't see why an association with gambling is bad, since that's exactly what you're doing here.
Also, it's not clear why freezing out 97% of the world is a bad idea either, if that 97% consists of people who break out in a cold sweat at the prospect of betting $100.
Last, what I'm recommending is just posted odds in different formats to make the props more accessible. I guess there's probably an argument for why increasing accessibility is a terrible idea, but off the top of my head I can't think of what it would be.
AFAIK there is a regulatory hurdle at the moment where prediction markets are essentially under the same umbrella as sports betting which has caused problems.
More generally I think gambling is broadly seen as at best frivolous, and at worst destructive and addictive. Whereas the vision of prediction markets (according to proponents at least) is that they can provide utility to society as a sort of truth oracle. Hence I see the association as unwanted.
Idk if gatekeeping normals would be seen as good or bad but, I don't think people familiar with moneyline odds are exactly the crowd the the prediction market wonks want to attract? But then IDK, might be interesting to pit academics and pundits against grizzled pragmatic gamblers, where everyone has skin in the game
This could all be 100% true, but if there's one 'prediction market' that's efficient (or close to it) it's the sports betting market, so barring some kind of keen insight on the part of the site developers I'd think the initial aim would be to use that model until evidence becomes available that it doesn't align well (for whatever reason) with these particular markets.
I don't know what the actual odds are that 'Joe Biden will blah blah blah by October 1st', but if I want to know what the actual chances are of Liverpool hoisting the UCL trophy in two weeks I'm probably going to a deep, liquid sportsbook with high house maximums like Pinnacle and seeing what they think.
I think in some sense it is easier for a sports betting market to become efficient, because it contains inherent uncertainty.
If you imagine a betting market on whether I will roll a fair six sided die and get a 1, the odds are going to quickly converge on the right answer. You don't need to research it in order to come up with a good sense of what the odds should be. In a market that's 100% luck, efficiency is easy to reach.
The more you remove luck from the market, the more useful research becomes. In soccer betting, if a goalie does really well in previous games, there are good reasons to expect that to be correlated with his performance next game.
But the questions that are most interesting, from a forecasting perspective, are questions where there's little inherent uncertainty, where there's no reference class to draw from, and different modelling assumptions lead to wildly different answers. Efficiency is hard to reach in that environment, but they're the questions that matter.
That's fine. My larger point was the posting odds in multiple formats doesn't have any obvious downside, and at least for me the current way that odds are posted (e.g. 'no $ .97') isn't intuitive at all. Maybe I'm not the kind of person these markets are trying to attract, but I know if I came across odds expressed in this manner I'd simply shrug my shoulders and drive on.
I might immediately understand "60%" to be the current probability estimate of the event in question, as provided by the wisdom of the market.
What I don't immediately understand (as someone who has never used a prediction market) is how that translates into the amounts of money I could win or lose if I were to enter the market now.
I'm good at mathematics/probability so I imagine I could work it out in about 30 seconds. Or I could read the FAQ.
But any prediction websites that wish to attract newbies by removing barriers to entry should consider taking account of this. Perhaps it's best to provide a site-wide preference, and somehow draw the attention of new website visitors to the feature.
Does 3/1 odds mean a probability of 1/4 or a probability of 1/3? (I believe both interpretations are used in different contexts, though maybe one of them is traditionally written with a colon instead of a slash?)
3/1 odds in a gambling context (used nearly universally for gambling odds in Britain and Ireland, and widely used for casino gambling and horse racing in the US) means that a successful $1 wager will pay out $3 in addition to you getting to keep your stake (or getting it back, if you paid in advance): you start out with $1 and end up with $4 for a $3 profit. Assuming the oddsmaker isn't taking a middleman's profit (which he almost certainly is), then that implies a 1 in 4 probability.
It just now occurred to me that one of the advantages of gambling odds formats, from the perspective of oddsmakers, is that it obfuscates the spread between exclusive options from which they make their profits. An options/futures market style odds format like that used by most current prediction markets makes the spread very explicit: if the bid on a "Smith wins the 2023 election" contract is 60¢ and the ask is 62¢, then it's pretty clear that there's a 2¢ spread between buying the contract (betting on Smith) or shorting it (betting against him). That spread is a lot less obvious if the odds for a bet on Smith are listed as -150 and the odds for a bet against him are listed at +163.
The 1 in 3 interpretation, I think comes from statistics, which talks directly about odds of a thing happiness rather than about bets and payouts.
I see that I was also thinking of the European "decimal odds" on which 3.0 means that a 1 Euro wager will return 3 Euros, for a profit of 2. It's confusing that European 3.0 means the same as British 2/1, but there's just going to be a systematic difference of 1 between the decimal and the fraction because one of them reports the payout and the other reports the profit.
Just a heads up-- I think what you're looking for here in the Smith example is -163 and +150; otherwise you have an arbitrage opportunity where you could profitably bet both sides.
I know this just an aside, and I'm honestly not trying to be annoying- just wanted to point it out.
Betfair Exchange, not an American market, has a UI that will translate the odds into whatever format you prefer: 10/1 (British odds) ,10% (probability), +900 (American odds), or 9.0 (their own native odds format).
If you are American, you'll need a VPN to access it, and you won't be able to deposit money. Note that Betfair is a regular sportsbook; you need to check Betfair Exchange to see the market version, and it is dominated by sports, not politics - and doesn't do markets on non-politics news events.
Anyway, decimal has one advantage for advanced gambling: you can get the decimal payout for a parlay (A and B and C) by just multiplying the payouts of the various events.
Right! I was actually thinking of Betfair last night when I was posting (or, if not Betfair, another UK-based exchange of one sort or another). I think Betfair used to take US action, but had to quit after the UIGA passed in 2006.
I seem to recall several betting sites having problems with people not liking/understanding the traditional American odds system when online betting was rolled out nationwide. I don't think it's actually a particularly popular or useful format.
If you have 100 99%-certain-Yes prediction markets, and one of them ends up resolving No, that's not a sign that prediction markets don't work, that's a sign that prediction markets work exactly like they should. One-in-a-hundred chances should happen roughly one time in a hundred.
It'd actually be a bad sign if every 99%-certain-Yes prediction market always resolved Yes.
For the record: data shows that "horses run to their odds", in that higher-odds horses win proportionally less often as the odds would indicate. The exceptions at both ends are called the "favorite/long-shot bias", in which people avoid very short-odds horses, such as a 1/5 horse (bet $2 to get back $2.40) because it's boring, and long-shots get more betting action than they should, because people love a big score. Also, in racing, the win/place/show takeout -- the house edge -- is 17% at most US tracks, which is a massive house edge for a gambling game, and that skews all the payouts low by that amount relative to real fair odds.
Maybe I've gotten too old, but I just can't figure out Discord servers for communities. I use them for voice and text chatting with small groups of friends, but every fan community Discord I've joined is unusable and baffling to me. How far back to you have to start reading to understand what in the world people are talking about? What's the point? What function does this serve?
It must serve a function, because they keep popping up everywhere, but it's not connecting to any of the functions I'm used to fan sites trying to serve. I'm probably just trying to make a square peg fit in a round hole. So what is the purpose of a Discord that is different from, say, a fan forum, wiki, or subreddit?
I think a forum is generally superior for the same purpose, and I think the driver is that a Discord server is easier to "run" than a forum (you don't have to figure out hosting, the forum software, etc etc).
The other justification I have seen offered is that the lower barrier of entry encourages more participation in a Discord server than an equivalent forum. I think this is broadly true, but this is really only worth the tradeoff for communities that are losing participation, which this one is not. I think the "more participation" part is short-sighted and comes at the cost of future community cohesion for the reasons you've listed.
My take is that the old community structures tended to have short-medium-long components, that is, day-to-day community chat for fun or whatever -> medium-term discussion that might be worth looking back at as reference -> mostly static info repositories. Common sets were stuff like irc-forum-wiki or ingame chat-forum-fansite and such, and Discord is a good fit for the short part, but I think people also think it can fulfill the forum function, which it really can't.
>One weird coda to this story: the process of choosing the new administrator continues, and (surprisingly!) two of the leading candidates are female. Someone might be about to make a lot of money.
My worry about prediction markets has always been, what if that person is the person that decides who the next administrator will be?
That's the type of manipulation I'm much more worried about, here.
I think the answer given is generally "so what?" If your goal of the market is to learn true facts, then the person making the decision betting on it will indeed be a really good way to learn true facts. They're exactly who you want to have betting, in fact. If, on the other hand, your goal is to make a speculative market where you can outwit the other traders and make a fortune, then it's bad but that seems like a less pro-social and more zero-sum kind of goal.
I think the counter-argument is that you don't want the decision maker to make the decision based off what will make them the most money, like betting on and then choosing the extreme underdog even though they're unqualified. And it seems hard to prevent that if you don't ban them from the market.
The trouble is that the information aspect spawns out of the "less pro-social and more zero-sum kind of goal." If there are not a lot of people who expect to make some money by being clever, the market doesn't work to give much information. No one bets on a prize fight where one of the fighters bets 1billion$ on himself to go down in the first round**. You don't get any information about who is the better boxer that day, just information around "what's this guy plan to do?" Which is not very interesting, particularly since the betting isn't supposed to be around what the guy plans to do.
** Unless the other fighter makes the same bet with reference to himself. Contest to see who goes down first?
>I think the counter-argument is that you don't want the decision maker to make the decision based off what will make them the most money,
Exactly.
If the person making the decision knows what they're going to do and bets on it and that make the market more accurate, that's good.
If the person making the decisions makes a different decision than they otherwise would have to make money off of the market, then that's bad, assuming there was a good reason for the decision they were making in the first place.
I'm reminded of the piece 'Markets are anti-inducitve'
You can't reliably know what a market is going to do in the future, because if that knowledge is public the market will shift to account for it, until you no longer know anything.
Prediction markets themselves being anti-inductive is good - that just means that all possible knowledge is accounted for in their predictions.
But what I worry about is that the ability of people responsible for predicted outcomes to make tons of money by hanging those outcomes will end up making *reality* anti-inductive.
If it's possible to make a good guess that something will happen in reality, and the current prediction market has converged on that price, then there's money to be made by making sure that thing *doesn't* happen.
People have motives to make sure reality reaches a point of being completely unpredictable and random, because they can always make money by forcing reality away from anything that can be predicted.
I think that would be a bad world to live in.
That's the extreme case, where prediction markets as big as financial markets and have as much power to shape society; before that point, it seems likely these effects will be limited in impact.
But it seems like advocates *want* prediction markets to get that big, and I'm still not sure what prevents these things from happening.
"the US government constantly disagrees with itself".
State governments, too. NY has a can deposit deposit of $0.05 so people don't just throw their cans on the side of the road. NY also has a prohibition on open alcoholic cans in an auto so people just throw their cans on the side of the road. It's not that the left hand doesn't know what the right hand is doing -- it's that the left hand doesn't *care* what the right hand is doing.
It probably had a bit more of an effect in the 70s and early 80s when most of the current beverage deposit laws were implemented. A 5¢ deposit in 1982, when NY's deposit law was passed, is equivalent to about 15¢ today.
I don't know how many cans get returned for their deposit, but a considerable percentage of the cans I picked up on Sunday were beer. Like most of them.
I actually think the point is not to prevent people from throwing cans on the side of the road - it's to incentivize extremely low-income people to walk down the side of the road picking up cans.
I agree with your larger point, but surely someone drinking a beer while driving assumes they aren't going to get pulled over? And therefore isn't worried about being seen with the empty can?
Anecdote: my father picks up cans once in a while for the deposit, and a decade and a half ago he was refused a crossing at the Canadian border in the wee hours of the morning because he had his regular trash bag of two dozen cans or so in the back seat, including some beer cans; they've held it against him up to the present day and we were questioned again about it on a crossing two years ago.
Anyway the point is I think there are reasons to not want to keep beer cans in your car other than being afraid of being pulled over, and not everyone is diligent enough to find the nearest bin.
That just sounds like the same sort of thing that's going on when I run the air conditioner in the summer and then proceed to cook dinner inside my house. I've got two things that I want to get done, and one of them has a side effect that the other one is designed to clean up.
>One weird coda to this story: the process of choosing the new administrator continues, and (surprisingly!) two of the leading candidates are female. Someone might be about to make a lot of money.
Is insider trading a risk with prediction markets (i.e, someone involved in picking the admin is making a huge bet)? On the one hand I would think not, since if someone moves the market to what they know will actually happen that improves the market's accuracy. On the other hand, every bet has two sides, and people will stop playing in the market if they know you can't win without inside knowledge of what will happen.
Somewhat off topic, but it's actually kind of funny that you're just finding out today that cloned dogs are available because I actually just found that out today too. Actually, with the recent Kentucky Derby, it got me wondering whether cloned race horses are a thing, and it turns out that they actually are- and may be priced pretty competitively with traditionally-bred race horses.
Cloning, albeit by splitting an embryo to make dozens of artifical IVF identical twins rather than taking genetic material from an adult animal, is also used in livestock farming.
1. I agree that sufficiently large prediction markets should be pretty accurate (at least on a certain type of question) and have the trust, aggregation, and clarity benefits you extol here. But we don't actually have any prediction markets like that. It's fine to play with the existing ones, and try to support their growth, but I think it's dangerous to index too hard on the specific predictions they make today.
2. In the case where some very rich person dumps a bunch of money into a large prediction market to manipulate it, and then a bunch of other people dump (in total) a similar amount of money into it to bring it back to a more accurate number, it seems like "a significant fraction of the economy is tied up in this one prediction" could cause problems, even though the market itself will have gone back to "normal".
One interesting thing about the ACX moderator market is that that's a market where someone can have private information that's *much* better than the public. So I'm wondering whether it was actually wrong to bet so heavily against the manipulator, given that for all the public knows they could have been the person who can choose the replacement.
Could Musk-owned Twitter host The Prediction Market That Actually Gains Traction - it will have the network, the high profile, the Musk genius. Makes sense to me. It could resolve the farce of Disinformation/Misinformation.
Probably not, they can't use real money any more than anyone else could. They could make really big play money markets, but I think Facebook already tried that. Given that Twitter is a bunch of small silos, I'm not sure how much that helps beyond Manifold.
Actually - this is something I worry a little bit about (in terms of prospects for Manifold the biz). I think Musk owning Twitter and making it a lot better would represent the most formidable threat to our company's earning potential.
But it's not something I worry a ton about - competition is a good thing! And if we're outcompeted because Twitter gets a lot better, I'd view our job as done, and we can go focus on something else (like building impact certs)
I wonder what an "Opinion and Debate" prediction market might look like. For instance, let's say you had a market in which the question is "Did Donald Trump inspire the Jan 6 attack on the Capitol?"
It's a matter of opinion not fact, but you could still have a betting market around it. The market could be a site in which public debate takes place over the question. The question could have a length of 3 months, after which, contracts close and payout and their current price. So such a market could close out at 53%, say, and that would be the payout.
Such markets would be open to manipulation, but so what? Would they be interesting markets? Would anyone participate in them? Would they mean anything at the end of the day?
I think this would basically be a prediction market for "what will the price of this prediction market be?" Since the market never resolves to 0 or 100, you have no reason to believe that your stocks will go up or down based on external facts. The only reason to think your stocks would go up is if you think other people believe it too. To turn a profit at 53%, you need to find someone who's willing to buy at 54%, and they need to find someone who's willing to buy at 55%, and so on and so forth.
But on the other hand, since the market resolves at the sale price, it would be very difficult to *lose* money on such a trade. If you buy at 53% and can't find a "greater fool," you have no reason to sell, you can just hold until the market resolves and lose nothing. So I guess prices would tend to rise to 100% over time? (Or get asymptotically close to it, since as prices rise it becomes harder to find a big enough fool to sell to.)
Point being, while people might buy stock in such a market for weird psychological reasons or because they think they can predict what their fellow traders think, there's no external economic incentive that ties the market to reality in any way. So I don't think it would tell you much about people's opinions.
I'm banning you for a day for kind of pointlessly insulting and psychoanalyzing me when you could have just asked.
I intended to discuss that prediction at the same time I discussed the fifty-odd other predictions in that post, but if it's that big a deal, I can explain my reasoning now. I think partly it was that the Supreme Court at the time had a 6-3 pro-choice majority (counting Roberts, who hates anything controversial) and I wasn't expecting two liberal justices in a row to die/retire and get replaced by conservatives. And everyone was predicting a "blue wave" in the upcoming midterms, which would have meant that even if two liberals had coincidentally died/retired it would have been hard for Trump to replace them with conservatives. And partly it was assuming that even if that did happen, it would be justices in the mold of Roberts who hate officially overturning precedent and would rather effectively declaw Roe while pretending it still stood.
There wasn't a real prediction market open in February 2018 when I made my prediction. One opened in July 2018 and said 25% chance by 2028. They had good reason to be higher than me since Kennedy had resigned in the meantime and they were going for 10 years instead of 5, but probably that wasn't enough to explain the full difference, and probably I had been wrong even with the information I had in February. I blame my overly simple understanding of how much the Supreme Court respects past precedent.
I apologize for the pointless psychoanalysing and maybe this response will go up in a month.
But let me explain my reasoning.
The whole idea of an ideological Turing test is useful here.
When you wrote the Silicon Valley party story it was funny, but the strawman twitter employee upset by Elon Musk taking over Twitter seemed to show a lack of understanding of the actual worries of the people who are upset. You have been amazingly insightful about memetics and how information ecologies are not our friends.
This goes to Steve Bannon and his "flood the zone with shit" idea. It takes advantage of the fact that these new information ecologies are a sort of wild west where a lie doesn't just get half way around the world before the truth ties its shoelaces, it gets half way around the solar system, sets up a Venus colony and starts nuking the truth from orbit. The blocking of anti vax stuff, or the Hunter Biden laptop story wasn't about "unpersoning conservatives," it was about some people at Twitter trying to make their little information ecosystem a bit less cancerous. It's not their fault that a disproportionate amount of prominent conservatives were posting anti vax stuff. (The Hunter Biden Laptop story would even have gotten actual traction at the time had Giuliani just let the WSJ verify it... which he didn't do.)
We are going to watch the son of the former dictator of the Philippines win an election today because they used Twitter, Facebook, chain whatsapp messages etc. to discredit his opponent.
Looking at the entire history of the Federalist society and the 50 year process to get a properly "Originalist" supreme court, not predicting the end of Roe when there are 5 hard federalist society Republicans is the action of Susan Collins.
Anyway, I would love to hear why you didn't predict the end of Roe in your 2021 or 2022 predictions (I didn't realize you hadn't updated the prediction in the past 4 years) and I apologize for coming off as snide.
I have been reading you for almost a decade and am a huge fan.
Of course, there are many terrible conservatives and conservative causes. But since Twitter, and the rest of the Woke Tech, go exclusively after them and are not only lax but often cheer for the progressive ones, their claims to be a force for good or whatever are laughable. They are culture warriors, pure and simple, and the sooner Musk shows them the door the better.
Given the general liberal proclivities of Scott's readership, I doubt many of his readers are anti-abortion, even the Republican ones. Most libertarians are pro-choice.
As someone who usually is very critical of rationalists at-large, I agree: there are a few high-profile commentators who are pro-life here (Deiseach, for example, and I think trebuchet as well), but to me they seem to be outliers. I think the "average ACT reader" would like Roe v. Wade repealed on states-rights grounds followed by all 50 states passing legislation that allows for some level of abortion.
Does the "average ACT reader" really believe in "states rights"? I know that many ACT readers believe in rights for many sorts of non-human entities, but states seem like a particularly unnatural fit.
They tend to prefer smaller, nimbler, more responsive governmental structures on many issues, but aren't outright anarchists. This is commonly referred to in the US as "state's rights" due to the Tenth Amendment being interpreted by them to reserve the bulk of civic administration to the states instead of the federal government.
I would use a somewhat different argument; I don't have strong opinions about exactly what level of government these decisions should be made on, but I do have strong opinions about the rule of law, which means the Constitution should either be followed or changed-then-followed on these matters.
Which, in practice, comes out to being "pro states rights" since the natural tendency of power is to centralise.
On "smaller, nimbler, more responsive" government, I think the US gets it wrong in many cases and devolves powers to an unreasonably low level, e.g. in giving local rather than state governments control over things like police and schools.
As a side counterpoint, local rather than state is most definitely the way to go in terms of control over schools or police. What you would want is a check or balance that is in the hands of the state - standardised assessments that let the local citizens know how their locally controlled school is performing vs others, for instance. We have the pendulum swung the other way in India, i.e lot more centralised/state control, and though admittedly the American govt is better than the Indian I've, the forces that centralisation puts into play are much tougher to handle.
I think it depends on definitions - a number of libertarian people are anti assault to the point of passing laws against it & getting the state involved.
More likely the strong pro abortion stance of the commentariat is a result of the underlying liberal/ left of center make up.
I'm pro abortion. Those fetuses know what they did :V
More seriously, though, I actually am pro abortion if the child would have a significant disability or if the parent is a teenager or otherwise not financially stable/not in a position to responsibly raise a child.
Except oddly, Ron Paul is anti-choice when it comes to abortion, and he's probably been the most popular Libertarian presidential candidate in my lifetime.
Well it comes down to your conception of human rights. If all humans have rights, then even a libertarian should be pro-life because abortion violates a prenatal human's right to life, and violates it without due process. I mean libertarians aren't anarchists, they do generally believe murder should be illegal. If you believe humans only have rights if they have certain qualities (consciousness, level of physical development, sentience, ability to have preferences, etc) and that prenatal humans lack the relevant qualities then naturally a libertarian would be pro-choice.
So libertarianism is not really that relevant a distincter for pro-life/prochoice, you have to know their conception of human rights and who they apply to.
It's a bit hard for me to interpret the answers on that poll. It sort of looks like the answers to that question add up to about 100%, but the answers don't look exclusive. I'm guessing anyone who checked one of those answers is interpreted as meaning that they think abortion should be legal in the case mentioned, but none of the others - but I see 9.5% checked "in case of incest", and I don't think I've *ever* heard anyone endorse that as the *only* case they'd allow abortion (particularly when severe disabilities of various sorts are other options on the list).
I think that's true, but only weakly so and I wouldn't count on it to predict the views of a non-random subset of libertarians. Never mind a group which isn't entirely and maybe not even mostly libertarian to begin with.
Libertarians are pro-most-choices, but they are as anti-killing and especialy anti-babykilling as pretty much the entire rest of the human race - "chosing" to kill a baby, even your own, is a wrongful act that is almost never justified. And whether or not one views an N-week-old fetus as a "baby" with a right to not be killed, is a question lots of people really sincerely disagree about and which is orthogonal to the question of personal liberty vs state power or whatever. Some libertarians do, some don't.
Mmm, it depends. The vast majority of libertarians believe murder-in-the-normal-sense should be illegal, but not all of us derive that from a right to life. There's a categorical-imperative argument that murder-in-the-normal-sense *must* be illegal; people don't want to be murdered, and adults and teenagers are quite capable of arming themselves and/or joining mutual-defence groups (and this isn't even criminal conspiracy if murder is permitted), so a society with legal murder would immediately turn into warlords in an arms race who themselves privately enforce "murder is banned" within their zones of control. Contradiction; QED. So even without actually believing in the right to life, you get "murder should be illegal".
But this argument has a *lot* of exceptional cases where allowing X form of killing doesn't result in the chaos and arms race that allowing general murder does. Duelling doesn't result in chaos, for instance, because someone can just not agree to duels if they don't want to be killed and thus there's no motive to start militias. Allowing kills in self-defence doesn't result in chaos, because someone can just not attack people if they don't want to be killed (with that said, you have to try to avoid cases in which two people can *each* legally kill the other in self-defence, otherwise you're back to at least a degree of arms race). And abortion (or even infanticide) doesn't result in an arms race because fetuses and infants aren't capable of fighting back either individually or collectively. You can to some degree even carve out an exception for killing slaves, so long as slave revolts are infeasible -history contains a few ugly examples of societies that did this and yet managed to endure for centuries, so as terrible as this kind of thing is it's clearly not incompatible with civilisation.
So if you're going absolute hardline "all is permitted that can be while avoiding 'bellum omnium contra omnes'", you get "abortion should be legal". On the other hand, not everyone who calls himself/herself a libertarian goes *that* far (at the very least, most balk at the idea of slavery).
"The future of abortion" sure is a little light on what overruling Roe v. Wade means for people and why we care. Would especially love an analysis on how federal abortion allowance would fit into the Archipelago and Atomic Communitarianism worldview.
re: misinformation, Prediction markets solve the problem of predicting what some authority will say at some future date, but not which authority to use for the settlement source. So it only works if the chosen authority eventually converges on the truth and actually reports it. Big media can't usually get away with lying about simple binary questions of fact, but if there's any ambiguity they can round it to whichever side they feel like and e.g. denounce the lab leak theory prematurely.
Funny enough, this project got started because of a short conversation I had with one of the founders (Austin) in the comments of a Mantic Monday post!
Scott Adams predicted specifically that Trump would win in 2016 *in a landslide*. He didn't outperform the prediction markets: his prediction was wrong.
Ditto to this. I'm already skeptical of proclaiming someone a prophet because of a single unlikely prediction (if you make 10,000 wild claims, one of them will probably come true even if the absolute odds for each one is 0.01%), but SA didn't even make a claim that came true! He made a claim that ended up being vaguely truth-adjacent! And I don't think his theory as to why (which boils down to "Donald J. Trump has powers of persuasion and charisma more properly belonging to a comic-book supervillain") reflects reality at all, either. As internal comms around 1/6 have shown, a lot of people in the GOP secretly hated him and only worked with him out of strict political necessity!
Nate Silver's prediction was a *lot* closer to correct than Scott Adams's - Silver was only off by a couple points in a couple states! The problem is that the effect of these differences is very discontinuous, and people are evaluating the correctness of the prediction by closeness in these effects, rather than in the thing itself.
The problem with prediction markets is that the best people won’t focus on events too far in the future. My friends are trading crypto futures. If they can’t make at least 30% pa on good enough volume, they won’t care.
Re. the theoretical Trump manipulation: skewing the prediction market with money might not work, because that can be viewed as an unexpected, unexplained deviation. But manufactured news in coordination with targeted media pieces, bold claims _and_ skewing the market with money might kind of work. Even more so, if a lot of people are watching the prediction markets as a gauge and trusting them, like the modern stock market.
Also, another problem which surely has been raised, but I'll do it again just in case. Insider trading, e.g. Scott buying tons of the "female moderator" option, and then appointing a female. I don't really see the solution to that. It can be outlawed, but will be extremely hard to police. I'm pretty sure this must have been discussed, so if someone could link me that, I'd be very grateful.
Insider trading is often considered to be a good feature of a prediction market - it means that the insider information gets out.
Of course, it's a bad thing when it *affects* the insider decision, rather than just being a way for the insider to cash in on insider information by releasing it through the market.
(For instance, selling shares in ImClone when you hear about the bad trial results is a way to make the price go down, so that more people know about the bad trial results even before the FDA rejects the drug.)
Doesn't the prediction of 10% chance of a female moderator strikes anybody else as just... wrong? I mean, if the user base is 90-10, making that prediction essentially boils down to believing that the next moderator will be chosen at random. That cannot possibly be true... right?
The fact that there are strong female candidates in contention (although we don't know among how many, I'd assume that it is less than 20) means that the probability was necessarily higher than 10%, and the secret better was more accurate than the market -- although I concede that 100% was obviously wrong as well.
Which means that the market going back down to 7% (lower than 10%!) is essentially saying that this prediction market, the one with the highest volume on the platform, is the least accurate (against the commonly held assumption that a higher volume will tend to higher precision). This seems to me the perfect case against prediction markets. Am I missing something obvious?
In every STEM company you've got a lot of men in the bottom layers, the first level of management and the top level. And a lot of women in HR and upper-middle management, i.e. admistrators. I don't think that's just random or sexism. They really like these jobs more.
So for scott's admistrator positions I would expect the density of women to be at least twice that of the surrounding milieu. So a market value of >20%
I give slightly higher than proportion odds to a female head mod for a slightly different reason. Being female protects you (somewhat) from a set of attacks. A bad actor won't blame sexism for their behavior being shut down, you're somewhat less of a natural target if the Eye of Sauron (woke media) decides to cast its gaze in our direction again, and some people will be a bit more cautious in their public attacks lest THEY come off as sexist. It's much the same reason that the WH press secretary in recent years is generally female.
There's a natural counter-argument that being female opens you up to additional sex-based attacks. Certainly the internet has plenty of people willing to toss incredibly offensive attacks at women. But I don't think this comes close to counter-balancing the protection, because those attacks are simply easier to deal with from a tactical perspective. If someone throws out a sexist attack on a female mod, you can ban them with >90% popular approval in this community. If someone accuses a male mod of sexism, you will not have near-unanimous popular approval to simply ban them.
I think 20% might be a better bet-line than 10% here.
It's a belief that the moderator will be chosen at random *with respect to gender.* It's probably going to be nonrandom on other attributes, like "how long have they been a member of the discord?" or "how frequently do they post on ACX?" If something has no bearing on your ability to be a discord admin, then the outcome *should* be random.
What this prediction is basically saying is "I have no special information on which of the candidates would make a better mod, but I believe the people choosing aren't sexist." That doesn't sound too surprising, if the candidates are all pretty well qualified. I'm guessing that "discord admin" isn't really the sort of job where you can be a rockstar who's 10x more effective than the next guy, so which candidate is "best" is going to be effectively random, or based on weird contingent facts like how much they vibe with the selection committee, which is something that you don't really have much information on.
"any more than the stock market is competing with investment banks and market analysts."
I thought they were competing-- that money flows to the more profitable markets. This is complicated by people wanting different things like risk vs. security or getting things wrong, or working on different time scales.
Do we need a theory for why there are different sorts of money markets?
The intended sense of "the stock market" here, I believe, is the *literal* stock market i.e. the physical and legal machinery allowing trade in securities, not the extremely-common metaphorical use to denote "the aggregated people who hold and trade securities".
Elon Musk made a fuss recently about index funds. Saying: if most of the market is index funds instead of people actually trading, then the market price records a lot of noise instead of an actual valuation of the various companies.
There's the same effect for prediction market vs experts. If you have a domain where the correct answer is found in the mouth of experts, then the prediction market is just the opinion of the expert plus noise. And the more people listen do markets instead of directly going to experts, the more noise I'm expecting to see in the markets.
Now I'm imagining a prediction market index fund: they buy a little of every opinion! Which, unlike the stock market, is sure to lose money because a prediction market is zero-sum. So really you should model a prediction market as being a bit like a stock market but with no index funds. No index funds means most traders will either be those with expert knowledge on the particular prediction, professional investors, and casual investors. Professional investors will have hedge funds and the like to be able to throw a lot of money at a market, and will have the resources and incentives to try to bet accurately. Casual investors will produce noise, but generally should be unable to counter professional investors with big pockets. Unless they band together somehow...the GameStop situation from last year does come to mind.
It doesn't strike me as particularly impressive that a prediction market can copy the performance of an expert? I mean, we already have the expert? The market adds nothing in this case.
I think it's likely that prediction markets are overrated, at least until you add insider information, but then that makes for a whole different problem instead.
How do you know which expert to listen too? How do *I* know which expert to listen too?
If everyone already agrees on who has the right idea, then yes, a prediction market doesn't add much. It seems to me that this is frequently not the case!
How does the prediction market know what expert to listen to, in that case?
I mean, *either* the expert is better and the market just copies (and with some delay, even), or the market actually performs better than any expert. In the first case, what's the point of the market in the first place?
Well, if there's anyone good at figuring out which expert to listen to, the prediction market will end up reflecting their views on which experts to listen too...
But the high-level meta-answer is that a prediction market performs boundedly-worse then any process anyone is actually using to formulate their beliefs, without you needing to know in advance to listen to that process and not others, because the best process will get higher weightings over time as people loose money to it.
I think this can sound unimpressive - "what, we only find out who was right answer once everyone else goes bankrupt?", but every trick you can think of to outperform it is also a trick the market can use.
This includes object-level processes like "look around you and try to figure it out", but also meta-level processes like "listen to people with degrees" and meta-meta-level processes like "listen to rationalists about which fields have experts worth listening too and how to find them", and even "infinitely-meta" processes like "notice when you are loosing money on prediction markets and change how you think".
If you, personally, already use the best possible approach to truth finding, you won't get anything out of checking a prediction market. And if everyone already knows the best possible approach to truth finding they won't either. But you probably don't, and they *definitely* don't.
But being better than simply the aggregate of all predictors doesn't sound all too hard? Sure, if an iterated market eventually kicked out all the people with poor judgement (by taking all their money, if nothing else), then it would slowly end up with a competent core that would create fairly strong predictions, but cases like sports betting or even election betting don't seem to bear that outcome out?
Looking at election betting in particular, it doesn't seem to be very good at establishing the *actual* chances of the candidates? We always get a lot of bonkers numbers, and Five Thirty Eight tends to scorn the "Scottish teens".
It doesn't even get you the performance of the best expert; it gives you the consensus of the top experts. Someone might statistically beat it out by an illegible margin.
But that's GREAT for any situation where it's hard or expensive for you to figure out what the actual consensus of experts is on a topic. Take... the Jones Act (which prohibits foreign-flagged vessels from doing interstate coastal trade.) Repealing that should decrease transportation costs (from first principles, increased supply), but unless you've done a deep dive on it you probably don't know whether that's likely to be a 1%, 10%, or 25% reduction in price. A prediction market lets you get a number for the expert consensus on it without spending the cost and effort of doing a deep dive into the field and the people in it yourself.
Is it actually the case that only the experts play? Otherwise, this could easily be interpreted as non-experts muddying the waters with their prejudices.
Sure, it might be irrational to, but it's not like people don't gamble at objectively poor odds.
In a sufficiently large market, there's too much money to be made. People looking to make money will come in and get rich off the amateurs (by correcting the market toward the expert positions.)
Note that this actually happened with the post-election Trump odds. A bunch of people thought there was a >15% chance that Trump would get inaugurated in 2021, even in early January. That market made a lot of people who paid attention to the experts instead of whatever propaganda a lot of money (but was distorted up by the popular amateur opinion.) As markets get larger, the cost to the amateurs to do the distortion goes up, and so the distortion should go down.
The amateurs lost about £40 million on BetFair Exchange on that exact market.
I think that was a case where there really just wasn't enough expert money to overcome the amateurs.
Note: it's not usually the actual experts investing their own money, but "smarts" - gamblers with deep pockets who make their money by betting with the experts against the amateurs.
Yep! Part of that is because prediction markets almost always have bet amount caps; if there are 5 smarts who know about prediction markets and want to make a bundle on correcting Trump inauguration odds, there's a limit to how much they can make (and consequentially how much they can correct the market.) Broader penetration of prediction markets (so there are more smarts), and raising that individual cap should both improve fidelity of the prediction market in cases like this.
BetFair Exchange is a huge gambling market with no caps - they do not have a general liquidity problem. There is a million pounds bet already on the 2024 Presidential election, and that's people locking money up for more than two years. The Australian federal election, in only a couple of weeks, has £2.5MM on the main market, and a few hundred thousand more on individual electorates and some other markets.
They will routinely have a million or ten bet on Premier League football match, and their record single bet is well into seven figure.
There just was so much money coming from Trump supporters after he had already lost that the smarts just couldn't soak it all up in 2020.
It was a liquidity problem, but a really exceptional one. Worth pointing out that the market ended up with well over half a billion pounds bet in total in 2020; this is why there just wasn't enough money to make the market.
Isn't that excessive chance a sign of failure in the market anyway, though? If the idea is that the market should be highly correct, then it *failed* here. That experts (or people trusting the experts) could make money off of it just shows that we should listen to the experts, not to the (clearly unreliable) market? If you trusted the market rather than the experts, you would *lose*.
Right. I don't want to use the boolean "is/isn't a failure" model, but it's certainly true that prediction markets were inaccurately high on Trump's inauguration in Dec 2020. They performed worse than a consensus of experts.
But at no point here have I claimed that prediction markets beat experts. The trouble is that finding who the experts are on a question isn't a free action. It's legitimately HARD for an outsider to figure out who the experts are in a given field, who the cranks are, and consolidate the generally nuanced and broad-scope expert opinions into a specific probabilistic prediction. This can all be done, and if you do it you'll beat prediction markets and can make money on them. (And in making money on them of course, you bring them closer to expert consensus.)
Put another way, prediction markets aren't competing with legitimate experts in the field. They're competing with the Washington Post or Fox News. They're incredibly legible and easy to find summaries of opinion on a topic, biased to more strongly reflect the opinions of those who the 'smart money' trusts (See Richard Gadsden's post immediately above yours here.) The argument in favor of prediction markets isn't that they're a magical truth-finding tool; it's that dedicated gamblers who are making money off the markets are more reliable at finding and presenting accurate predictions than journalists. It really makes sense from an incentives point of view. Journalists lose money when they lose attention. The gamblers lose money when they're wrong.
Indeed, arguably the 'failure' in 2020 was mostly due to a lack of smart money in the system. Almost all prediction markets have features like betting caps which prevent the gamblers from making more than X dollars each on a given market (and so limit the force they can apply.)
> The argument in favor of prediction markets isn't that they're a magical truth-finding tool; it's that dedicated gamblers who are making money off the markets are more reliable at finding and presenting accurate predictions
Right, so ignore the prediction-market odds unless you want to know where the average gambling addict gets their news, but do try to find people with a proven track record of making money off prediction markets and then read their blogs or something.
Sure. Such a market would probably be set up in a form like "If the Jones Act is repealed before Jan 1, 2023, will interstate transport costs (by X standard metric) drop by at least 10% from their 2021 values to 2023 values?" and resolve inconclusive if the act isn't actually repealed.
Presumably it would only be set up if there was significant interest in getting it repealed, and so a decent chance of resolving conclusively.
But then we wouldn't know until afterwards if the market had good predictions, and once we know, it would no longer be useful. It seems that we would only be able to find the prediction market reliable for similar, repeated events (such as perhaps sports), where a level of consistency and credibility could be established?
The idea is that prediction markets can be validated as a class; if they've worked with reasonable fidelity to expert opinion on questions structurally similar to a repeal of the Jones Act, then we can expect them to have such fidelity in that case as well.
As someone who has been betting on political markets on BetFair Exchange (effectively a prediction market) for nearly 20 years, I'll say a couple of things about market manipulation.
First, someone spending a lot of money on a market to manipulate the price can do so, but they need to keep pumping in money to keep the price up, and the market will attract more and more "smart" money so it gets increasingly expensive to do so. This was done, intentionally, by one of the campaigns for the leader of the Liberal Democrats in 2006, and estimates were that they spent about half a million (sterling), and ran out of money about two weeks before votes were cast. This did, however, probably elevate them as the main challenger, though they still lost - but they would likely have finished third without this.
Second, the weight of each person's opinion depends on how much money they have; the 2020
US Presidential election market was not resolved until the electoral college met and voted (in December). Large numbers of Trump supporters kept the Trump probability around 10% from when the TV networks called the election until after. Like many people, I put in a few thousand to extract an easy 10% ROI, but the Trump supporters were putting in tens of millions and there just wasn't enough "smart" money to push the probabilities any lower.
It is possible that a really large amount of money can manipulate the market, either by a single person or by enough people that have been fooled by misinformation or disinformation. But it is rare and not having a restriction on insider trading (there's no legal restriction) means that the markets are probably more accurate. I know of many cases of political insiders making extremely healthy profits by betting on insider information about likely election results. Campaign staffers often use this to make up for relatively poor salaries.
My issue with prediction markets is that the arguments for them often don't distinguish between theoretical prediction markets with perfect liquidity and an infinite number of participants etc vs the prediction markets that are likely to exist in the real world. You can maybe get close to that with markets on extremely major issues like presidential elections, but that doesn't necessarily transfer to everything else. And proving how it would transfer tends to be elided
Sometimes I think that the best moment to get Transnistria back is now: when Russia is weak, busy, and generally opposed by everyone. After the war in Ukraine ends, if Moldova tries to get Transnistria back, Russia can pay full attention to the issue.
But then I remember that Ukraine is only able to fight Russia successfully because of all the foreign help. Moldova cannot expect to get a comparable amount of help, because the budgets are limited, and Ukraine is a more visible cause. So it could end up with Russia attacking Moldova and taking even more of their territory. For example, Russia could accept the loss against Ukraine, and negotiate something like "we will stop the war now and give you half of the conquered territory back, but you will stop helping Moldova".
Scott, I don't mean to accuse you of dishonesty, just a normal oversight, but the last line of the Discorcasting section seems very close to a "heads I win, tails you lose" situation. Are you saying now, before it's decided, that if a female moderator is selected it would be strong evidence *against* the accuracy of prediction markets? If not, I don't think we can take the low price as evidence for prediction markets' resistance to manipulation.
Question about prediction markets generally: is there a good explanation for how/why they will handle long-term, unlikely events?
For instance, PredictIt offers shares that Bernie Sanders will *not* be the 2024 dem presidential nominee for $0.97. Payout is $1.00, so ~3% increase. Let's say I don't think Bernie has a shot -- I could buy a ton of shares, but I'm effectively locking my money up for *two years* for a 3% return. That's barely better than treasury bond yields over 2 years, and there are probably better places to park my money.
For longer-term predictions with even longer shots, this problem seems even bigger. It seems like prediction markets will necessarily be pretty off for these kinds of predictions, where they cap out at maybe 20:1 odds. Does that sound right? Are there other factors I'm not thinking about that will keep them accurate for this?
> Disclaimer that Manifold probably can’t handle probabilities this small correctly and there’s no reason to think 0.2% is more realistic than 2%. It’s not 10% though.
Is there any reason to think 0.2% is more realistic than 0.002% ?
I feel like it's very hard to stick reasonable probabilities on these tails. The question becomes "What's the probability that my mental model of social reality is totally wrong in some fundamental way?" and I think that's very hard to answer sensibly.
I don't understand why you don't cover sports betting, Scott. Sports books are mature prediction markets with lots and lots of money sloshing around in them. Are they good? I dunno, I haven't looked into it, I'm in the camp that all this is kinda dumb. But it seems like a very good arena in which to test your claims for prediction markets. For instance, "public teams" (teams that have chronically inaccurate lines because people like to bet on them independent of their quality; the Yankees, Packers, and whatever team LeBron is on at the moment are public teams) are a notorious facet of sports betting that seem pretty bad for your claims about prediction markets. But I dunno, maybe public teams are a myth, or they were a thing in the 90s that got driven out by the rise of quantitatively sophisticated professional gamblers.
I know you find sports gauche, but if you don't care enough about prediction markets to overcome that, I don't see why I should be asked to care about them either.
Prediction markets kind of remind me of usury. Usury, of course, being the practice of lending money at interest, which was widely regarded in the western world as a sin for over a thousand years. The sin, very generally speaking, was loan sharks: someone loans you money at interest, you can't pay it back, you're in debt to them forever. It was also seen as cheating people: you (the lender) don't actually do any work, so why should you get paid? Today that objection sounds silly: they get paid to compensate the risk of not being paid back, and loans provide a lot of social good so what's wrong with making a profit? But back then all they had as their model of loaning money at interest was loan sharks, so it was condemned as a sinful practice. Today the world economy runs off of loaning money at interest, and the idea of banning it would be almost unthinkable.
So to with prediction markets. Gambling is considered a sin, and for good reason. We all know people who have ruined themselves through gambling, and it's hard to see what social benefit gambling provides anyone besides entertainment. It makes sense that just about every religion and most respectable secular authorities condemn gambling as evil. Yet I can imagine a future, a few hundred years from now, where the idea that people once considered betting on the outcome of events as a terrible sin will be just as surprising as when someone today finds out that lending money at interest was once considered a terrible sin. It's not that our values will change, but the form of the thing itself will change. Just as the free market transforms avarice into societal wealth generation, prediction markets could transform gambling into a prosocial good as well.
> it's hard to see what social benefit gambling provides anyone besides entertainment
Well, this is different with prediction markets. The benefit is having good information, and the possibility to tell people who talk nonsense to either bet on their beliefs or shut up.
Oh I agree. I think that "gambling" now has the same social connotation as "usury" did to a medieval priest. If prediction markets take off then they will do to gambling what fractional reserve banking did to usury.
Gambling a sin? I grew up playing cherry bells and turkey bingo at my (Catholic) church. In high school we had to sell raffle tickets as a fundraiser for the school. Sit down at a poker table in a casino and you'll see a wide range of religious backgrounds represented.
> A lot of times, people ask: what if some very rich person tried to manipulate a prediction market? Like, what if there was a market on whether Trump would win the presidency, and Trump himself dropped $100 million into making it say YES, either out of vanity or in the hopes that people would think he was “inevitable” and stop resisting.
> And the theory has always been: if traders see a clearly mispriced market, they’ll rush to it and correct the mispricing in order to make easy money. Trump would be out $100 million and the market would stay correct.
> We’ve never been able to test the theory, and until someone drops $100 million into a real money market, we never will.
Increases in life expectancy were largely coming not from old people living to older ages, but from young people not dying young, in particular not as infants. It looks like this is something we will reverse - old people are likely going to live to be older. Unfortunately, I do not expect I will benefit from these advances, I was born in the early 80s, but I do think that most kids born these days will live much longer lives than we see today
Well it’s both reducing infant deaths and increasing adult longevity. I think we’ve hit a wall on the latter. We were prolonging sickly old age but even that has reversed. We are not really solving the problems of ageing and some of it looks really bad to me. As in looking at my parents generation (not my actual parents luckily) there’s a lot of people in very bad shape. My age group is looking after them and most of us don’t want to go through that, I think euthanasia will increase in popularity.
Another edition of "please someone explain betting markets to me." In this edition, I made a bet on Manifold and am super confused by the result.
I think there's a solid chance (maybe 20%?) that Republicans will make Abortion illegal nationally within the next 5 years. I saw that the Manifold market was at 6% with $198 bet, so I decided to weigh in. I submitted $25 for a Yes bet.
Two confusing things happened: First, I received 111 shares of "Yes." I would expect 416 shares ($25 divided by $0.06 per share). Why did the shares cost $0.23 instead of $0.06?
Secondly, the Manifold probability jumped from 6% to 78%! That's a huge jump! One, it's above my own true estimate, which is dumb. Two, my $25 bet is about 1/9th of the $223 total market value, so why is my Yes bet outweighing everyone else? Looking at the graph, all the other bets (totaling $198) moved the probability up or down by a couple percent. What's going on?
Guessing: the price of $0.06 was the lowest price anybody was willing to sell for, but there weren't 416 shares of that. So first you cleared the 6-cent shares, then your offer went on to eat up shares that were more and more expensive, until you had $25 of shares. The last of those shares were at $0.78, so that's the new "market price".
Did anyone sign up for the forecasting tournament mentioned in the previous open thread? If so, did you get a confirmation email after submitting the signup form?
This may be a pointless bit of language carping, but I wonder if it's worth considering a slight change in language, away from terms like wrong and right, when you make probabilistic predictions.
"By the way, in 2018, I got this horribly wrong - I said only a 1% chance of Roe v. Wade getting repealed in the next 5 years."
It's not obvious to me that this prediction was wrong! I mean, I understand this use of the term, but strictly speaking, the claim that you made has not been invalidated or disproven by subsequent events. And part of the value of prediction markets is precisely to push people away from simplistic right/wrong pigeonholing of views. One way to do that is to not use the simplistic language when discussing them. Instead you could use something like, "This now looks like a poor prediction; I lost that bet; in hindsight, it looks like I may have evaluated some factors incorrectly." The potential downside is that you sound a bit less human/normal, of course, so it's bit of a trade off.
It's an interesting question whether a probabilistic prediction could ever be wrong. As a dumb example, if you flip a (ordinary) coin and predict a 1% chance that it will land heads, then your prediction is "wrong" in the sense that it's working from an incorrect model of how the world works.
I think it's fair to say that Scott's 2018 prediction was wrong in this sense, that he was working from an incorrect model of how the world works, and I think that most expert SCOTUS-watchers would have been able to tell him that at the time. It wasn't a combination of unlikely events that got us from 2018 to this point, it was a sequence of fairly likely ones.
Thanks, yeah, I take the point about getting the percentages wrong. I'd like to push back on it a bit! I haven't delved into the details yet, but if Scott's 5-year prediction was made at the beginning of 2018, then he would have been right if RvW wasn't repealed before the end of 2022. The ruling is due to come out in June, so his prediction held up for 4 and a half years of the five year period. And there's a lot of happenstance and coincidence that can affect any court case, so the chances of the ruling coming 8 months later are good.
Of course, this may be one of those cases where I'm applying one of Scott's heuristics that work 99% of the time... I'm not sure!
> On the DSL threat, someone brought up that the US government was quicker to call the Russian invasion of Ukraine than most prediction markets. I agree this happened. In some sense, it’s unsurprising; the US government has spy satellites, moles in the Kremlin, and lots of highly-paid analysts. Of course they should do better than everyone else.
Given that I seem to have a pretty good track record of prediction, it's a shame that I've been banned from both Astral Codex Ten as well as the Motte. The consistent pattern there seems to be small-minded mods telling me that I'm not NICE enough, but as we learned from the Ukraine invasion, reality is not always NICE. My embrace of reality in preference to NICENESS could have averted a war and a pandemic, but because people don't like taking advice from somebody who isn't NICE, I guess we needed to have a whole bunch of people die instead. Maybe the community should stop optimizing for niceness? It doesn't seem to be a very useful (or rational) strategy.
Please tell how you could have averted a war and a pandemic. Congratulations for predicting it in advance, but how would you have used those predictions to change the outcome, and how did getting banned stop your plans?
I mean, "immediately ban all transit from China, from anyone else with cases, or from anyone who doesn't do the first two" would have fairly-reliably prevented X non-China country from getting CoViD (and in a lot of cases, the public will was there and this was sheer betrayal/incompetence by politicians). And I think he means that "if people had done what I wanted, they would have averted this", which is a whole lot weaker than "if I hadn't been banned from a subReddit, we would have averted this".
"My embrace of reality in preference to NICENESS could have averted a war and a pandemic, but because people don't like taking advice from somebody who isn't NICE, I guess we needed to have a whole bunch of people die instead"
Wow, Mr. Really Big Important Guy, I had no idea you were Emperor of the World! Please tell me why your subordinates in major world governments refused your lawful instructions and why you did not eviscerate them all as punishment?
Or contrariwise, you might be Just Some Guy with an opinion, like millions of others, and why should anyone listen to you claiming you are right, over the other guy claiming he is right that Covid is caused by fluoride in the water?
Oh, you published an e-book? Like hundreds of thousands of other people. Yes, a self-published book by Literally Who has immense credibility.
Make a prediction, right here and right now, that can be checked. Don't do the "I hinted at this in my book" after the fact bollocks, that's astrology. Make a definite prediction that X will happen by time Y, and then we can see if you were right or not.
Get a majority of administrator candidates on board. Buy. Candidates announce they now identify as female. Sell. Candidates announce they now identify as male. Buy...
(Admittedly this works better if the market is real money and you can offer them a percentage of the profits. But I personally would try to get in on the first gender market manipulation scheme just for the anecdote.)
If they already have an opinion on the topic, including the probability estimate, it is almost *free* money. Even people who do not care about money too much would probably take free money.
And even if they do not care about free money either, as long as they tell their opinion to someone else, that other person can place the bet for themselves.
I have not read all the comments / posts / etc related to prediction markets, so apologies if this is redundant/covered. But I have not seen it covered in what I have read.
Prediction markets seem neat now, because they seem to be high-signal. People mostly seem to want to predict, not manipulate. People mostly want to be accurate according to factual reality.
This seems to mirror the early Internet. Very less overall content but higher average quality. But then came mass adoption, Facebook, Twitter. More recently, concerted, sustained disinformation.
I would be concerned what happens if prediction markets become less niche, more noticed: "successful" by some metrics. Say Russia or the CCP or some other government-level entity with enormous resources takes an interest in manipulation. The goal here might not be to drive the outcome, merely to muddy the waters and maximize noise. So you could not simply get rich by disagreeing.
on Xi re-elected CCP General Secretary: as somebody who bought No on Polymarket, I see the main possibility to be him retaining power but with a different title. Think of Putin's 'castling' with Medvedev as a comparison. It's not probable, but IMO it's over 10%
Why do you think markets didn't even react to the news of Putin's health? I personally think that's the biggest predictor he might stop being president
Aubrey de Grey says that there are seven (and only seven) things that make us age, and he has solutions (for some understanding of "solution") for all seven.
So
(a) do you trust Aubrey de Grey? Do you trust him more or less after seeing a picture of him? Opinions, as they say, vary.
(b) do you trust that his solutions will actually work (I'm assuming the next hundred years or so are long enough to iron out any engineering issues; the conceptual issues are the point)
(c) [and this is the big one] do you believe that "society" will allow these solutions to be tested then implemented? I flip back and forth on this. On the one hand, rich people and other countries are going to be happy to let this go where it goes. On the other hand, "society" has a long history of being fairly successful in slowing down or suppressing certain types of medical research, especially stuff that requires large scale effort not just a mad scientist in lab (eg much interventionist genetic medicine). So...
> general sentiment
Are prediction markets not just another form of echo chamber? Is there some obviously-true analysis of that somewhere?
The fact that you end up losing all of your money if you're wrong, and end up earning a bunch of money if you're right, and that we can look at the histories of traders and prediction markets, kind of makes me less concerned about this
I think it would be hard to make a sock puppet into a mod, since a user generally needs to have a well-established history in the community to get trusted with a mod position, and an account that was created right when the previous admin decided to step down wouldn't have much history.
(And if you've been posting frequently for years on an account to create the appearance of a well-established user, that's less a sockpuppet and more a main account.)
Can you give an example?
I agree that if, eg, Trump found a way to hack the election machinery to ensure he'd win, he would win. And if at the same time he bets on himself winning in a prediction market, he could make money off of it. That seems like generic insider trading. Is that your point, or are you trying to say something more complicated?
as you can see from the OP, attempted manipulation can make a minor market the most popular market on the site due to all the people taking the manipulator's money. There's no market small enough for continuous manipulation to be cheap.
What if Trump bet $1B that he would lose the election, then resigned?
What if the person choosing the next moderator was the one who poured $3500 into the market predicting it would be a woman, then chose a woman?
What if a scientist is very close to discovering a cure for cancer and their efforts are highly publicized, and they notice the markets predict they will publish a cure in 6 months, so they spend all their money on 2-year predictions and then chill in Tahiti for a year and a half while people keep dying of cancer?
Etc.
In a sense, yes, this is the same as a CEO shorting their own company and then intentionally crashing it.
But I *think* it's already illegal for a CEO to do that? Or at least they would never work again? Whereas I don't see any penalty for doing that in prediction markets as they currently exist.
I think you're generally correct that, for whatever worry we could come up with about manipulating prediction markets and creating bad outcomes, there's some analogy you could draw to the same type of behavior in a stock market.
But I think stock markets are incredibly highly regulated in all sorts of ways in order to prevent those examples, *and also* stock market manipulation *does* happen and *does* lead to all kinds of really bad outcomes, fairly often.
It’s a good point. Presumably the SEC (or an equivalent entity) would eventually get involved and apply stock market-like insider-trading laws to these markets if they got big enough?
Ie you can only trade on public information, cannot directly manipulate the markets, etc. It seems that the same “fairness” intuitions we apply to stocks would apply here.
Which would be a shame since insider trading helps price discovery. In the Trump example above, the $1B entering the market is the smart/correct money...
But that is only true because insider trading is possible. In the Trump example, if insider trading were illegal, presumably he would try to win.
I'm not sure this is the case. In the worst case we're talking about people changing their behavior after bidding on an unlikely outcome they can affect, not just insiders betting on outcomes that they have private information on.
For example I think the Trump imaginary scenario was sketching the idea that Trump was not planning to resign, but decided that a 10x yield (or whatever) on $1B was enough to convince him to do so.
I don't think the SEC would actually need to get involved. One high profile case of insider betting here would make a lot of potential bettors really paranoid about how the markets are working and who is betting what. (Who in the "Did one guy make those bets? I should look into that." sense, not the literal identity sense.) My guess is that such outcomes would become pretty self limiting as users dig into what's going on during the event, and the problems would be discovered and fixed before the regulators recognized a problem after the event and investigated.
If the insider was betting on what he planned to do, it actually would give good information, too. If someone wants to spend $10,000,000 to tell me ahead of time that they are going to tank a company, ok?
On the other hand, if they are trying to manipulate the information output because they want the company to tank based on misperceptions of its viability, that might be a problem. Maybe not, though, as other people who know about the company's viability would bet against them hard, which improves the information, and perception only goes so far towards affecting reality.
Edit: forgot the Trump thing. You might see some regulation around people starting the markets not being able to bet on the outcomes, or something like that. Possibly that would be imposed by the market makers themselves. I am given to understand that it is highly frowned upon for athletes to bet against themselves in contests, for example, though I don't have any idea how true or enforced that is.
Insider trading laws aren't about fairness, they're about theft. If you trade on misappropriated information (e.g., things you learn in your capacity as a CEO *belong* to the shareholders, and trading on it for your own benefit deprives them of value), that's a crime; if you learn something no one else knows through legitimate means, it's arguably unfair to other traders who don't know it but it's not a crime.
For two examples mentioned in this thread, if the person responsible for choosing the next moderator for the Discord was the one manipulating the market, it would make a ton of sense, and a make him a ton of 'cash' (though you've said elsewhere he claims it isn't him). For another example, in the question about whether the identity of the person who leaked the Roe v Wade docs will be known before 2023, if the 'No' side has a significantly large enough pot, the leaker could buy a ton of 'Yes' and make a killing by outing themselves.
For a bit more morbid example, I don't remember who said it first, but basically any prediction market is also a de facto assasination market. Buy enough 'No' positions on a 'will this person still be in power' contract, and you can make a killing, making a killing.
That's part of the point of a prediction market. If someone is trying to covertly manipulate the actual event, then someone who knows about it is likely to reveal this manipulation by buying up the market.
Looking at the first graph of the PredictIt market, 4% for the decision in Dobbs still seems awfully high. That's not based on any actual knowledge of the case or the court. It's based on the fact that "MI" stands for Michigan, not Mississippi.
1. Haha, very funny.
2. I just checked the PredictIt page and it says "MS". I think the caption is courtesy of Scott and he flubbed it because he was in a rush.
3. PredictIt markets have official rules sections more detailed than the market title, so even if the title said MI the rules section would probably say the right state.
Not to say that PredictIt markets never end up resolving in a way contrary to the expectations of many market participants! I can think of a solid handful, all a lot of fun.
Do tell!
That's my fault, not PredictIt's - the original caption wouldn't fit in the space provided so I rewrote it and messed it up.
I figured - if only making money on prediction markets were that simple...
Missed a main point: $14000 can turn a completely tertiary question into the biggest market. Or in other words, these prediction markets are eminently manipulable for minimal cost. This is a recipe for f***ery.
So a manipulator should dump a lot of money into a market they don't care about so people don't notice movements in a market they do care about?
I would think this generally allows someone to "make a mountain out of an anthill". The manipulator would probably lose money on the market, but can gain something (awareness/money/distraction) outside of the market. E.g. highlight a vague, but possible inoffensive point of a some law and make it into a national issue worth debating about during the election cycle. It's already possible, of course, with traditional media, astro-turfing on the net, etc., but prediction markets probably would offer yet another channel of attack on the public discourse. The market probably should correct such attacks quickly enough, though, unless they're backed by a cult-like movement which can sustainably bet along with the instigator, however irrationally (e.g. a TV preacher congregation willing to throw money wherever their leader points to).
But such attacks could only work reliably on small-scale markets, I think. Otherwise it would take too much money to make even a blip against really hot predictions.
They could also just spend the money directly advertising their cause, lol.
And be lost in the general din of advertisement. Of course, the effectiveness of this tactic would depend a lot on popularity of predictions markets and the audience that will participate and track them.
Anything can be used to make money; a "market" manipulable with small amounts of money is a pre-condition to scams.
Manifold is a play money market where nobody is expected to spend real cash, which is why someone actually spending that much made everything crazy. The point I draw is "even if one person has much more money than any other individual and spends it on manipulation, everyone else will gang up to de-manipulate it"
> Manifold is a play money market where nobody is expected to spend real cash
That's obviously the correct response here, but I've never quite understood why you don't think it undermines this whole enterprise. You seem very interested in the play money markets, at least partly as a source of information about how large-scale real-money markets could work. But surely the differences are too large to draw many serious conclusions about one from observing the other? If you've written about this in detail somewhere I'd be interested to read it.
I mostly do think it undermines the enterprise. I don't think it makes it 100% useless, but it makes it not great for anything with real stakes. I do think this current lesson still holds because it's a case where the play-money-ness ought to work against rather than for it, but it still succeeded.
I think the real issue is that right now, these are mostly a playground for making money or messing around.
The more these are used as real indicators of important events, the less useful they will become, because they will be more prone to both manipulation for the purpose of external manipulation, as well as because there will be more less informed people involved.
This is an issue with the real stock market - market manipulation can greatly manipulate the value of certain targeted stocks (like Gamestop or Tesla) but it's more difficult to manipulate the value of all stocks in general. However, if whatever the stock in question is happens to be important at that time, then the manipulated stock might be a big deal.
Also, in the stock market, most forms of market manipulation are illegal and the SEC pursues such cases vigorously (perhaps too vigorously?).
And then Scott complains that the SEC is making it too hard to set up unregulated prediction markets for real money.
Given the importance of the stock market, I think the SEC being harsh about it is warranted.
Honestly I think that Musk belongs in prison for some of the things he's pulled.
I envision a scenario of government fueled propaganda working in concert with prediction market manipulation (assuming the markets become relevant in the future).
So Putin (or whoever) says something about his plans, and has his propaganda machine back it up, and simultaneously has a hidden organization bid up the price as if what Putin is saying is correct. There's enough noise available to the public that it's harder to make a prediction, and the market itself seems to be backing up a particular conclusion, making it less likely people bet against it.
That assumes that the information isn't actually available to prove the case either way, but with sophisticated propaganda, that's often the case. In the specific instance of Trump bidding up the "Trump Wins" category with $100 million - right now people would be fools to bet $100 million against Trump winning in 2024. Is he definitely going to win? No, but there's a strong enough chance that it would be extremely dumb to bet that much money against it. Maybe the true percent should be 40%, but because of the size of the bets in favor, the market says 60%. So Trump's $100 million keeps the market floating, enhancing his own chances of actually winning (people like winners).
What if the manipulator were a bit more subtle? Instead of dumping thousands of dollars at once and shifting the market straight to 100%, you'd want to shift it slowly to avoid attracting attention to the market.
I think this is a highly atypical market anyway since it's too prone to insider trading. Without knowing anything about how the moderator is being selected, I would assume that anyone placing significant bets far away from the consensus is probably party to some inside information.
I think it doesn't matter how fast you go, if you see a market at 90% when you think it should be 10% you're going to take notice.
But doesn't a sudden market movement like that make people who *don't* know it should be at 10% take notice? Like yeah, the ones who have familiarity with and grasp of the subject matter would have noticed either way. But surely there are a whole host of people who'd simply have noticed that this random market was suddenly the biggest in the history of the site, and had jumped suddenly from one extreme to the other, and realised what had happened (or at least deemed it worthy of inquiry)?
Whereas if you do it more subtly and slowly, none of those people need be alerted.
That's what seems really weird to me. Insider trading is generally a *good* feature on prediction markets, because it gets you a *lot* more accuracy than just trading on general knowledge. In fact, getting insiders to reveal their private information is sometimes said to be one of the mean benefits of prediction markets.
If the market reacts this strongly to block someone spending a lot of money on it, wouldn't it also react this strongly to some insider trying to cash in on their knowledge? And wouldn't that defeat the purpose of the market?
If the insider trader has good intel and the market "fights" them on it, then it's the insider who gets to make a bunch of money when the question finally resolves.
The same could have been said for cryptocurrency (play money where nobody expected to spend real cash), so not so clear to me that your point is as strong as you evidently believe it is.
How are they analogous? From the beginning, cryptocurrencies were intended to have real-world value. I don't think anyone expects their prediction-market play money to ever be exchangeable for goods or services or dollars.
Cryptocurrencies were intended to play a real world role - that of currency.
In actual reality, cryptocurrencies were used initially primarily to evade capital controls plus facilitate cyber crime.
Prediction markets are intended to fulfill a real world role: prediction.
In reality, how will this be skewed by the unscrupulous and/or conniving?
This will be any place where prediction confers advantage.
Election expected outcome is a known influencer of turnout.
Conferring of credibility onto otherwise utter nonsense endeavors is another.
I'm am 100% certain there is an infinite portfolio of these since we already know the widespread impact of past prediction purveyors such as astrologists and palm readers.
I never participated in the "what do you most disagree with Scott about?" thread, and I agree with your points regarding why prediction markets seem useful. But, I think if there is something I do often disagree about, it's the usefulness of measures for turning prediction markets into *drivers* of policy. I think there are some cases where this is probably a good idea, but I think that in a lot of cases, attempting to make the step from using them as an analytical tool to using them as a policymaking tool runs headlong into Goodhart's Law.
Well said, that’s a situation where there would be clear incentives to manipulate the market in advantageous ways - like cutting out the middleman of having to bribe tons of congressmen with donations. I don’t see a way they can make policy that avoids this unless it’s on very narrow interests without finanical stake - but then what’s the point of such a specific tool?
I don’t know much about this and would be interested to read a steelman by Scott or another
So long as the mispricing is more than 5%+10%/year a lot of people will gang up to un-manipulate it. This happens regardless of the motivations of the manipulator, so I don't think it makes any difference whether the manipulator is motivated by the fact that the prices will be used for policymaking.
Doesn't it depend on how many big money people care about the result of the manipulation more than the potential profit they can make?
if they have more money than everybody else combined and really really want to throw it away manipulating a market, ok, but manipulation on that scale would be pretty obvious and people wouldn't trust the probabilities.
Will it really be so easy to identify manipulation? For any market where it's even plausible that someone has important non-public information, how can even the smart money tell the difference between 'person with inside knowledge places large, accurate bet' and 'manipulator places large, inaccurate bet'?
Unusually high volume, and common knowledge among the high ranking forecasters that it seems insanely mispriced. It wouldn't go unnoticed, but it could be hard to distinguish insider trading vs inefficiency vs manipulation. People would take the numbers with a grain of salt even if they can't determine which.
This reminds me of the betting market for who would get the iron throne. Enough insiders bet on Bran early to make him a -1000 favorite before the show aired, which looked crazy to anyone without the insider knowledge.
If you're asking a some forecasting experts whether the market should to be trusted or is mispriced, why bother with the market at all? You could just ask the experts what price they think it ought to be.
In this case the experts are determined objectively by their track records on the prediction market, instead of whichever clown the fourth generation of Sulzbergers wants to give an NYT column to.
I mostly agree with this - I'm hopeful that one day prediction markets could be involved in policy, but this would be after we understand them a lot better and have much more experience with them.
https://www.metaculus.com/questions/9751/futarchy-adopted-by-2070/
"once they became important enough to matter, biased interests would take them over"
Once [prediction markets] became important enough to matter, [the US government] would take them over.
What do they get from taking it over? Unlike a pundit, you can't just choose to make the prediction market say anything you want (unless you make it stop being a functional prediction market). If it says a number that's wrong, people will make money by betting against it.
"make it stop being a functional prediction market" would probably be the whole point.
That's not so much "taking it over" as "destroying it". It would be like seizing control of a newspaper just so you could shut it down.
Certainly that's a thing that could happen, but it's extremely different from corrupting the newspaper to print favorable stories. It's not the same problem that Scott was talking about in that sentence you quoted.
(You can't _pretend_ that the prediction market is still functional, the way you could pretend that a newspaper is still impartial, because regular people can place bets in a prediction market and then make money if they're right. If you just made your web site claim the consensus was 90% when it was actually 10%, you'd get a ton of people betting on 10%. Then when the thing doesn't happen, you've either got to pay off all those bettors--in which case this wasn't too different from just placing a large bet on an honest market--or else refuse to pay them, in which case all the bettors will realize your so-called prediction market is a fraud.)
Imagine the US government creating thousands of sophisticated sock accounts with virtually unlimited funds. Then they place normal looking bets in a consistent direction, thoroughly pushing markets in a particular direction. At the same time, they release particularized subsets of real information at their disposal, creating an official source of data that backs up what they are betting on.
Is there any reason that they couldn't do this for things that are false as much as for things that are true? Sure, the US government might not want to blow millions of dollars on every question that comes up, but do you think a political party wouldn't be willing to spend huge amounts of money when they are in charge and the results of prediction markets have substantial affects? (Imagine a prediction market for abortion, with the outcome affecting a national law for or against). Bonus, if you can actually swing the outcome the way you want, you get your opponent's money as well as your own back!
That was supposed to be the advantage of doing it through crypto like Augur, but that seems to have died down for reasons I've never heard.
Is this true? Has the US government taken over the NYSE or NASDAQ, in the sense that if a business they support's stock goes down, they can force NYSE/NASDAQ to make the stock go back up again?
I am a layman on the specific subject but that does not seem like a fair comparison; I would think that lack of explicit stock market manipulation is not because there are no avenues or desire for the US government to do so (simple imaginary path: well-targeted tightening or loosening of regulations) but rather because there's recognition that short-term stock price manipulation probably doesn't have all that much bearing on the long-term performance of whatever business the company actually wants to engage in.
There may be some analogous reason to not attempt manipulation of prediction markets but I don't see any at the moment; if they become sufficiently important, the current prediction itself exerts some non-total but non-negligible effect on opinions (whether of predictors or not), and it would seem useful to put a thumb on the scale in some cases. I don't think you can make a safe assumption that practical prediction markets will be immune to internal corruption and infiltration and the like, which will open up means of manipulation that are not possible within the regular rules of prediction markets.
Hmm, rereading that I recognize that there's a similar argument for putting a thumb on the scale for stock prices, but I would again attribute it to an understanding that it's not worth the squeeze, rather than not having any utility, and I would maintain that the cost and difficulty of manipulating a prediction market are likely to be lower than manipulating a major stock exchange.
What about all the "too big to fail" banks and the money dumped into supporting them? That may not be direct manipulation but neither is it letting the market sort itself out.
What's the analogy here? The government might pass a law causing the prediction to be true/false? But that's just how things normally work, the prediction market isn't involved at all
I think the analogy is that the government officials derived some personal utility by not letting the banks fail, and so used their official powers to make that happen. The government officials didn't force NASDAQ or DOW etc. to just write the stock prices higher, but they did force us all to pay a crap ton of money to make the stock prices higher via bail outs.
In terms of prediction markets, if the outcome of the market "Is policy X a good idea?" goes down, are officials who want X going to have an incentive to either spend money to make the market go towards YES, or perhaps more subtly, promise to change or actually change other rules such that X seems like a good idea, even if the rules changes are net negative? It might well devolve into a "I am going to break things until people agree that what I want is a good idea" situation. Admittedly, a lot of policy is like that, a series of kludges and band-aids on a system instead of ripping out or fixing the broken bits, but then that's why I can easily imagine a government manipulating a prediction market that way.
At the extreme, if prediction markets become important enough to drive policy, the government will start a "Board of Analytic Predictive Services" to control the spread of prediction market manipulation. Markets that the government deems manipulated, by virtue of making obviously false predictions in the eyes of the board members, will get BAPed until they display the right prediction.
Or more extreme, the government will just make their own market, make all others illegal, and then have a pricing algorithm such that the predictions are remarkably close to suggesting what the government officials wanted to do anyway.
Have you noticed that you can't buy futures contracts on onions? XD
Why go through a middleman in this way? Obama engineered the saving of GM.
This doesn't appear to be a good analogy, unless prediction markets could be circumvented to achieve the same policy goals (which is obviously the case right now in 2022, but we're talking about a future in which prediction markets serve some major purpose in society) and the government chose to go through prediction markets anyway.
Why don't these markets post odds in the traditional American odds format? Seems like either a silly oversight, or maybe it's just needless exclusionary. If these sites are interested in getting more traffic and action, then a first step would be post odds in a format that the typical gambler can understand.; i.e., if something has a 60% of happening then it's -150, or if it has a 20% chance of happening it's +400.
I think a tiny fraction of Americans understand American moneyline odds. Every newspaper article called Rich Strike an 80-1 long shot, not a +8000 one. (And that's not even getting into the fact that overseas people probably wouldn't be familiar with American odds either.) Most Americans understand what 60% means (or at least I'd like to think so).
Came here to say this. I've never even *seen* the "traditional American odds format", and I'm American.
I'd never heard of those, but this article is a handy short explanation of the different systems:
https://www.investopedia.com/articles/investing/042115/betting-basics-fractional-decimal-american-moneyline-odds.asp
I suppose it depends what you are accustomed to using; I'm used to the 'fractional odds' so the -200 and +600 don't make sense to me.
Thanks. I'm British and grew up with the 'fractional odds' system, without even realising it had a name. It never occurred to me that other countries would use other systems.
Given all the Americans that are replying here saying they've never heard of the Money Line odds system, perhaps it shouldn't be referred to as "the American odds" system. (A bit like how 10^12 should probably no longer ever be referred to as a "British billion".)
Am American, this is the first time I've ever encountered this odds format. From your comment alone I can make no sense of it.
I've also never gambled, except for casually (as in "Bet you $10 you can't make that basketball shot" with my cousin sort of thing). What percentage of Americans gamble enough to be familiar with this kind of odds format?
I don't think they intend for their primary market to be regular non-prediction-market gamblers, and I think you might be suffering from some social myopia on how widely understood the traditional odds format is. I would wager (ha!) that among the general population, people who both understand it and prefer it to a % chance are a small minority; I would even make a weak wager that with the recent expansions in casual "nontraditional" gambling that people who understand and prefer this format are a minority even among people who gamble.
Probably because the designers and their expected early adopters weren't/aren't sports gamblers. The current format is transparent both to people who are familiar with basic statistics (contract price trivially translates to percentage probability) and people who are familiar with futures and options (you're spending the posted contract price to buy a thing that will have different values depending on what winds up happening. I had no idea how to read American sports betting odds until I looked it up just now, and I suspect the same is true of most other non-gamblers.
Hi Eric,
Just a couple of points:
The current format, at least to me, is indecipherable. Taking the Ethereum market, for example-- I have no idea what 'no $.97' means, in terms of the risk/reward ratio. I'm not saying that the current layout isn't intuitive, only that at least to me it isn't. I think I can assume that the 'no' is the chalk, but that's as far as I can get.
Also- I'm sure you're right re: the intended audience, but if one of the goals is to create action it seems to me like it would make sense to offer odds in a few different formats that gamblers can understand. There are thousands of people out there who would happily take a $2000 flyer on a random prop, just for the hell of it, and the chances of drawing those people in would (I think) improve considerably if the odds were posted in a more conventional format. Most sportsbooks post odds in both American odds (for Americans) and decimal odds (for Europeans) for that reason, and doing something similar here would probably increase the liquidity of these props.
I have no familiarity with regular gambling nor prediction markets, but looking at the Polymarket format, I noted that it presents "No $0.97" in opposition to "Yes $0.03", adding up to $1, and I intuitively understood it to mean that the cost of a single unit of a wager (I don't know what the correct term for this is) on "no" is $0.97 and for "yes" is $0.03, and either would pay out $1 if it won.
Since you again casually used a term that only a tiny fraction of people would understand ("chalk"), I would guess that you are part of a demographic likely to interpret these presented formats in a very different way than the average layman, and I do think the prediction markets have aimed their odds presentations more at the average layman than at experienced gamblers.
Ha! That's fascinating. Thanks for sharing. So when you saw the layout it was clear that you were laying 97 cents to win a dollar on 'no', or laying 3 cents to win a dollar on 'yes'? I've looked at that graph for 15 minutes and that never occurred to me, though now of course that makes perfect sense.
To be fair it's not like I've never seen this format before in passing, but today is probably the first time I actually tried to interpret it. I don't know that it's *the* most intuitive format but I think it's at least a bit more newbie-friendly than the regular American odds format, for which I'd be in the same situation as you - I'd take forever to figure it out without looking it up.
I can totally understand that, now that you've brought it up. And there does seem to be something about betting odds specifically that makes them exceptionally difficult to interpret in unfamiliar formats, even when the format in question has been explained to us (I don't know if this is true for other numerical representations, though it seems likely that it is). For example, my wife-- who is not a gambler-- finds European decimal odds intuitively easy to understand, while I still struggle with them after twenty years ( I have to convert them into a percentage, and then convert them into American odds, before I have a clear picture of what I'm being asked to bet on).
Bad optics (association with gambling is bad) and also hell no that is horrible scheme for 97% of the world. Not trying to be mean but this idea is so bad I can't tell if it's a joke or not
I don't see why an association with gambling is bad, since that's exactly what you're doing here.
Also, it's not clear why freezing out 97% of the world is a bad idea either, if that 97% consists of people who break out in a cold sweat at the prospect of betting $100.
Last, what I'm recommending is just posted odds in different formats to make the props more accessible. I guess there's probably an argument for why increasing accessibility is a terrible idea, but off the top of my head I can't think of what it would be.
AFAIK there is a regulatory hurdle at the moment where prediction markets are essentially under the same umbrella as sports betting which has caused problems.
More generally I think gambling is broadly seen as at best frivolous, and at worst destructive and addictive. Whereas the vision of prediction markets (according to proponents at least) is that they can provide utility to society as a sort of truth oracle. Hence I see the association as unwanted.
Idk if gatekeeping normals would be seen as good or bad but, I don't think people familiar with moneyline odds are exactly the crowd the the prediction market wonks want to attract? But then IDK, might be interesting to pit academics and pundits against grizzled pragmatic gamblers, where everyone has skin in the game
This could all be 100% true, but if there's one 'prediction market' that's efficient (or close to it) it's the sports betting market, so barring some kind of keen insight on the part of the site developers I'd think the initial aim would be to use that model until evidence becomes available that it doesn't align well (for whatever reason) with these particular markets.
I don't know what the actual odds are that 'Joe Biden will blah blah blah by October 1st', but if I want to know what the actual chances are of Liverpool hoisting the UCL trophy in two weeks I'm probably going to a deep, liquid sportsbook with high house maximums like Pinnacle and seeing what they think.
I think in some sense it is easier for a sports betting market to become efficient, because it contains inherent uncertainty.
If you imagine a betting market on whether I will roll a fair six sided die and get a 1, the odds are going to quickly converge on the right answer. You don't need to research it in order to come up with a good sense of what the odds should be. In a market that's 100% luck, efficiency is easy to reach.
The more you remove luck from the market, the more useful research becomes. In soccer betting, if a goalie does really well in previous games, there are good reasons to expect that to be correlated with his performance next game.
But the questions that are most interesting, from a forecasting perspective, are questions where there's little inherent uncertainty, where there's no reference class to draw from, and different modelling assumptions lead to wildly different answers. Efficiency is hard to reach in that environment, but they're the questions that matter.
I have absolutely no intuition for what -150 or +400 mean, but I know exactly what a 20% chance is.
I can never figure out that odds format, and I'm grateful they use normal numbers like "60%" which everyone can understand immediately.
That's fine. My larger point was the posting odds in multiple formats doesn't have any obvious downside, and at least for me the current way that odds are posted (e.g. 'no $ .97') isn't intuitive at all. Maybe I'm not the kind of person these markets are trying to attract, but I know if I came across odds expressed in this manner I'd simply shrug my shoulders and drive on.
I might immediately understand "60%" to be the current probability estimate of the event in question, as provided by the wisdom of the market.
What I don't immediately understand (as someone who has never used a prediction market) is how that translates into the amounts of money I could win or lose if I were to enter the market now.
I'm good at mathematics/probability so I imagine I could work it out in about 30 seconds. Or I could read the FAQ.
But any prediction websites that wish to attract newbies by removing barriers to entry should consider taking account of this. Perhaps it's best to provide a site-wide preference, and somehow draw the attention of new website visitors to the feature.
E.g.:
- "No odds (just percentages)"
- "Also include fractional odds"
- "Also include money line odds"
- "Also include decimal odds"
I'd give you 3/1 odds that the "typical gambler" would be more familiar with international formats than the American one.
Does 3/1 odds mean a probability of 1/4 or a probability of 1/3? (I believe both interpretations are used in different contexts, though maybe one of them is traditionally written with a colon instead of a slash?)
3/1 odds in a gambling context (used nearly universally for gambling odds in Britain and Ireland, and widely used for casino gambling and horse racing in the US) means that a successful $1 wager will pay out $3 in addition to you getting to keep your stake (or getting it back, if you paid in advance): you start out with $1 and end up with $4 for a $3 profit. Assuming the oddsmaker isn't taking a middleman's profit (which he almost certainly is), then that implies a 1 in 4 probability.
It just now occurred to me that one of the advantages of gambling odds formats, from the perspective of oddsmakers, is that it obfuscates the spread between exclusive options from which they make their profits. An options/futures market style odds format like that used by most current prediction markets makes the spread very explicit: if the bid on a "Smith wins the 2023 election" contract is 60¢ and the ask is 62¢, then it's pretty clear that there's a 2¢ spread between buying the contract (betting on Smith) or shorting it (betting against him). That spread is a lot less obvious if the odds for a bet on Smith are listed as -150 and the odds for a bet against him are listed at +163.
The 1 in 3 interpretation, I think comes from statistics, which talks directly about odds of a thing happiness rather than about bets and payouts.
I see that I was also thinking of the European "decimal odds" on which 3.0 means that a 1 Euro wager will return 3 Euros, for a profit of 2. It's confusing that European 3.0 means the same as British 2/1, but there's just going to be a systematic difference of 1 between the decimal and the fraction because one of them reports the payout and the other reports the profit.
https://www.investopedia.com/articles/investing/042115/betting-basics-fractional-decimal-american-moneyline-odds.asp
Just a heads up-- I think what you're looking for here in the Smith example is -163 and +150; otherwise you have an arbitrage opportunity where you could profitably bet both sides.
I know this just an aside, and I'm honestly not trying to be annoying- just wanted to point it out.
Betfair Exchange, not an American market, has a UI that will translate the odds into whatever format you prefer: 10/1 (British odds) ,10% (probability), +900 (American odds), or 9.0 (their own native odds format).
If you are American, you'll need a VPN to access it, and you won't be able to deposit money. Note that Betfair is a regular sportsbook; you need to check Betfair Exchange to see the market version, and it is dominated by sports, not politics - and doesn't do markets on non-politics news events.
Odds format converters are everywhere on the net, such as https://www.aceodds.com/bet-calculator/odds-converter.html
Anyway, decimal has one advantage for advanced gambling: you can get the decimal payout for a parlay (A and B and C) by just multiplying the payouts of the various events.
Sure, but being able to set a setting on your account and get whatever format you prefer is still better than having to convert.
Right! I was actually thinking of Betfair last night when I was posting (or, if not Betfair, another UK-based exchange of one sort or another). I think Betfair used to take US action, but had to quit after the UIGA passed in 2006.
Yeah, they've been huge for a long time, but they are definitely coming from the sportsbook end, not the prediction market end.
I seem to recall several betting sites having problems with people not liking/understanding the traditional American odds system when online betting was rolled out nationwide. I don't think it's actually a particularly popular or useful format.
Doesn't Rich Strike winning the Kentucky Derby prove prediction markets don't work??
If you have 100 99%-certain-Yes prediction markets, and one of them ends up resolving No, that's not a sign that prediction markets don't work, that's a sign that prediction markets work exactly like they should. One-in-a-hundred chances should happen roughly one time in a hundred.
It'd actually be a bad sign if every 99%-certain-Yes prediction market always resolved Yes.
It proves that they do work - you'd expect a horse with 80:1 odds to win, on average, once every 80 years.
For the record: data shows that "horses run to their odds", in that higher-odds horses win proportionally less often as the odds would indicate. The exceptions at both ends are called the "favorite/long-shot bias", in which people avoid very short-odds horses, such as a 1/5 horse (bet $2 to get back $2.40) because it's boring, and long-shots get more betting action than they should, because people love a big score. Also, in racing, the win/place/show takeout -- the house edge -- is 17% at most US tracks, which is a massive house edge for a gambling game, and that skews all the payouts low by that amount relative to real fair odds.
I knew that it wasn't strictly true - but it is closer to the truth than "this long shot win proves odds aren't real"
TIL of the ACX discord server. Anybody have an invite link, please?
https://discord.gg/RTKtdut
(this link was also posted in the recent open thread)
Thanks!
Maybe I've gotten too old, but I just can't figure out Discord servers for communities. I use them for voice and text chatting with small groups of friends, but every fan community Discord I've joined is unusable and baffling to me. How far back to you have to start reading to understand what in the world people are talking about? What's the point? What function does this serve?
It must serve a function, because they keep popping up everywhere, but it's not connecting to any of the functions I'm used to fan sites trying to serve. I'm probably just trying to make a square peg fit in a round hole. So what is the purpose of a Discord that is different from, say, a fan forum, wiki, or subreddit?
I think a forum is generally superior for the same purpose, and I think the driver is that a Discord server is easier to "run" than a forum (you don't have to figure out hosting, the forum software, etc etc).
The other justification I have seen offered is that the lower barrier of entry encourages more participation in a Discord server than an equivalent forum. I think this is broadly true, but this is really only worth the tradeoff for communities that are losing participation, which this one is not. I think the "more participation" part is short-sighted and comes at the cost of future community cohesion for the reasons you've listed.
My take is that the old community structures tended to have short-medium-long components, that is, day-to-day community chat for fun or whatever -> medium-term discussion that might be worth looking back at as reference -> mostly static info repositories. Common sets were stuff like irc-forum-wiki or ingame chat-forum-fansite and such, and Discord is a good fit for the short part, but I think people also think it can fulfill the forum function, which it really can't.
Typo thread:
I think “predicting the results of the recent Supreme Court *link*.” was meant to be “predicting the results of the recent Supreme Court *leak*.”
"DSL threat" -> "DSL thread"
>One weird coda to this story: the process of choosing the new administrator continues, and (surprisingly!) two of the leading candidates are female. Someone might be about to make a lot of money.
My worry about prediction markets has always been, what if that person is the person that decides who the next administrator will be?
That's the type of manipulation I'm much more worried about, here.
I think the answer given is generally "so what?" If your goal of the market is to learn true facts, then the person making the decision betting on it will indeed be a really good way to learn true facts. They're exactly who you want to have betting, in fact. If, on the other hand, your goal is to make a speculative market where you can outwit the other traders and make a fortune, then it's bad but that seems like a less pro-social and more zero-sum kind of goal.
I think the counter-argument is that you don't want the decision maker to make the decision based off what will make them the most money, like betting on and then choosing the extreme underdog even though they're unqualified. And it seems hard to prevent that if you don't ban them from the market.
The trouble is that the information aspect spawns out of the "less pro-social and more zero-sum kind of goal." If there are not a lot of people who expect to make some money by being clever, the market doesn't work to give much information. No one bets on a prize fight where one of the fighters bets 1billion$ on himself to go down in the first round**. You don't get any information about who is the better boxer that day, just information around "what's this guy plan to do?" Which is not very interesting, particularly since the betting isn't supposed to be around what the guy plans to do.
** Unless the other fighter makes the same bet with reference to himself. Contest to see who goes down first?
>I think the counter-argument is that you don't want the decision maker to make the decision based off what will make them the most money,
Exactly.
If the person making the decision knows what they're going to do and bets on it and that make the market more accurate, that's good.
If the person making the decisions makes a different decision than they otherwise would have to make money off of the market, then that's bad, assuming there was a good reason for the decision they were making in the first place.
I'm reminded of the piece 'Markets are anti-inducitve'
>https://www.lesswrong.com/posts/h24JGbmweNpWZfBkM/markets-are-anti-inductive
You can't reliably know what a market is going to do in the future, because if that knowledge is public the market will shift to account for it, until you no longer know anything.
Prediction markets themselves being anti-inductive is good - that just means that all possible knowledge is accounted for in their predictions.
But what I worry about is that the ability of people responsible for predicted outcomes to make tons of money by hanging those outcomes will end up making *reality* anti-inductive.
If it's possible to make a good guess that something will happen in reality, and the current prediction market has converged on that price, then there's money to be made by making sure that thing *doesn't* happen.
People have motives to make sure reality reaches a point of being completely unpredictable and random, because they can always make money by forcing reality away from anything that can be predicted.
I think that would be a bad world to live in.
That's the extreme case, where prediction markets as big as financial markets and have as much power to shape society; before that point, it seems likely these effects will be limited in impact.
But it seems like advocates *want* prediction markets to get that big, and I'm still not sure what prevents these things from happening.
"the US government constantly disagrees with itself".
State governments, too. NY has a can deposit deposit of $0.05 so people don't just throw their cans on the side of the road. NY also has a prohibition on open alcoholic cans in an auto so people just throw their cans on the side of the road. It's not that the left hand doesn't know what the right hand is doing -- it's that the left hand doesn't *care* what the right hand is doing.
Realistically, I don't think the deposit meaningfully affects how many cans get thrown onto the side of the road.
It probably had a bit more of an effect in the 70s and early 80s when most of the current beverage deposit laws were implemented. A 5¢ deposit in 1982, when NY's deposit law was passed, is equivalent to about 15¢ today.
I don't know how many cans get returned for their deposit, but a considerable percentage of the cans I picked up on Sunday were beer. Like most of them.
I actually think the point is not to prevent people from throwing cans on the side of the road - it's to incentivize extremely low-income people to walk down the side of the road picking up cans.
+1
I agree with your larger point, but surely someone drinking a beer while driving assumes they aren't going to get pulled over? And therefore isn't worried about being seen with the empty can?
Anecdote: my father picks up cans once in a while for the deposit, and a decade and a half ago he was refused a crossing at the Canadian border in the wee hours of the morning because he had his regular trash bag of two dozen cans or so in the back seat, including some beer cans; they've held it against him up to the present day and we were questioned again about it on a crossing two years ago.
Anyway the point is I think there are reasons to not want to keep beer cans in your car other than being afraid of being pulled over, and not everyone is diligent enough to find the nearest bin.
That just sounds like the same sort of thing that's going on when I run the air conditioner in the summer and then proceed to cook dinner inside my house. I've got two things that I want to get done, and one of them has a side effect that the other one is designed to clean up.
>One weird coda to this story: the process of choosing the new administrator continues, and (surprisingly!) two of the leading candidates are female. Someone might be about to make a lot of money.
Is insider trading a risk with prediction markets (i.e, someone involved in picking the admin is making a huge bet)? On the one hand I would think not, since if someone moves the market to what they know will actually happen that improves the market's accuracy. On the other hand, every bet has two sides, and people will stop playing in the market if they know you can't win without inside knowledge of what will happen.
> One weird coda to this story: the process of choosing the new administrator continues, and (surprisingly!) two of the leading candidates are female.
Two out of how many?
Less than 18.
Somewhat off topic, but it's actually kind of funny that you're just finding out today that cloned dogs are available because I actually just found that out today too. Actually, with the recent Kentucky Derby, it got me wondering whether cloned race horses are a thing, and it turns out that they actually are- and may be priced pretty competitively with traditionally-bred race horses.
Here's the article I read (though I cannot vouch for the trustworthiness of the publication), please remove if links aren't allowed: https://robbreport.com/lifestyle/sports-leisure/cloned-horses-competition-2881549/
Cloned horses have been commonplace in polo for some time now: https://www.vanityfair.com/news/2015/07/polo-horse-cloning-adolfo-cambiaso
Cloning, albeit by splitting an embryo to make dozens of artifical IVF identical twins rather than taking genetic material from an adult animal, is also used in livestock farming.
1. I agree that sufficiently large prediction markets should be pretty accurate (at least on a certain type of question) and have the trust, aggregation, and clarity benefits you extol here. But we don't actually have any prediction markets like that. It's fine to play with the existing ones, and try to support their growth, but I think it's dangerous to index too hard on the specific predictions they make today.
2. In the case where some very rich person dumps a bunch of money into a large prediction market to manipulate it, and then a bunch of other people dump (in total) a similar amount of money into it to bring it back to a more accurate number, it seems like "a significant fraction of the economy is tied up in this one prediction" could cause problems, even though the market itself will have gone back to "normal".
One interesting thing about the ACX moderator market is that that's a market where someone can have private information that's *much* better than the public. So I'm wondering whether it was actually wrong to bet so heavily against the manipulator, given that for all the public knows they could have been the person who can choose the replacement.
IIRC Skeward (the guy who chooses the replacement) said it wasn't him.
TBH, that name doesn't exactly sound reassuring for someone who is claiming not to skew a market.
nominative determinism strikes again
Yeah, but he could have been lying, or there could have been someone else with insider information somehow.
To be clear I'm not saying that's the case, I'm saying it should have been a plausibly hypothesis with the information people had at the time.
Scott, I'd be interested to hear any thoughts you might have about the ideas in this post: https://www.lesswrong.com/posts/GNJNJ2cRyvSLp2owH/speculations-concerning-the-first-free-ish-prediction-market
It's a year old, but I stand by it (aside from the Kalshi optimism, which turned out to be misplaced).
Could Musk-owned Twitter host The Prediction Market That Actually Gains Traction - it will have the network, the high profile, the Musk genius. Makes sense to me. It could resolve the farce of Disinformation/Misinformation.
Probably not, they can't use real money any more than anyone else could. They could make really big play money markets, but I think Facebook already tried that. Given that Twitter is a bunch of small silos, I'm not sure how much that helps beyond Manifold.
Actually - this is something I worry a little bit about (in terms of prospects for Manifold the biz). I think Musk owning Twitter and making it a lot better would represent the most formidable threat to our company's earning potential.
But it's not something I worry a ton about - competition is a good thing! And if we're outcompeted because Twitter gets a lot better, I'd view our job as done, and we can go focus on something else (like building impact certs)
I wonder what an "Opinion and Debate" prediction market might look like. For instance, let's say you had a market in which the question is "Did Donald Trump inspire the Jan 6 attack on the Capitol?"
It's a matter of opinion not fact, but you could still have a betting market around it. The market could be a site in which public debate takes place over the question. The question could have a length of 3 months, after which, contracts close and payout and their current price. So such a market could close out at 53%, say, and that would be the payout.
Such markets would be open to manipulation, but so what? Would they be interesting markets? Would anyone participate in them? Would they mean anything at the end of the day?
I think this would basically be a prediction market for "what will the price of this prediction market be?" Since the market never resolves to 0 or 100, you have no reason to believe that your stocks will go up or down based on external facts. The only reason to think your stocks would go up is if you think other people believe it too. To turn a profit at 53%, you need to find someone who's willing to buy at 54%, and they need to find someone who's willing to buy at 55%, and so on and so forth.
But on the other hand, since the market resolves at the sale price, it would be very difficult to *lose* money on such a trade. If you buy at 53% and can't find a "greater fool," you have no reason to sell, you can just hold until the market resolves and lose nothing. So I guess prices would tend to rise to 100% over time? (Or get asymptotically close to it, since as prices rise it becomes harder to find a big enough fool to sell to.)
Point being, while people might buy stock in such a market for weird psychological reasons or because they think they can predict what their fellow traders think, there's no external economic incentive that ties the market to reality in any way. So I don't think it would tell you much about people's opinions.
I'm banning you for a day for kind of pointlessly insulting and psychoanalyzing me when you could have just asked.
I intended to discuss that prediction at the same time I discussed the fifty-odd other predictions in that post, but if it's that big a deal, I can explain my reasoning now. I think partly it was that the Supreme Court at the time had a 6-3 pro-choice majority (counting Roberts, who hates anything controversial) and I wasn't expecting two liberal justices in a row to die/retire and get replaced by conservatives. And everyone was predicting a "blue wave" in the upcoming midterms, which would have meant that even if two liberals had coincidentally died/retired it would have been hard for Trump to replace them with conservatives. And partly it was assuming that even if that did happen, it would be justices in the mold of Roberts who hate officially overturning precedent and would rather effectively declaw Roe while pretending it still stood.
There wasn't a real prediction market open in February 2018 when I made my prediction. One opened in July 2018 and said 25% chance by 2028. They had good reason to be higher than me since Kennedy had resigned in the meantime and they were going for 10 years instead of 5, but probably that wasn't enough to explain the full difference, and probably I had been wrong even with the information I had in February. I blame my overly simple understanding of how much the Supreme Court respects past precedent.
I apologize for the pointless psychoanalysing and maybe this response will go up in a month.
But let me explain my reasoning.
The whole idea of an ideological Turing test is useful here.
When you wrote the Silicon Valley party story it was funny, but the strawman twitter employee upset by Elon Musk taking over Twitter seemed to show a lack of understanding of the actual worries of the people who are upset. You have been amazingly insightful about memetics and how information ecologies are not our friends.
This goes to Steve Bannon and his "flood the zone with shit" idea. It takes advantage of the fact that these new information ecologies are a sort of wild west where a lie doesn't just get half way around the world before the truth ties its shoelaces, it gets half way around the solar system, sets up a Venus colony and starts nuking the truth from orbit. The blocking of anti vax stuff, or the Hunter Biden laptop story wasn't about "unpersoning conservatives," it was about some people at Twitter trying to make their little information ecosystem a bit less cancerous. It's not their fault that a disproportionate amount of prominent conservatives were posting anti vax stuff. (The Hunter Biden Laptop story would even have gotten actual traction at the time had Giuliani just let the WSJ verify it... which he didn't do.)
We are going to watch the son of the former dictator of the Philippines win an election today because they used Twitter, Facebook, chain whatsapp messages etc. to discredit his opponent.
Looking at the entire history of the Federalist society and the 50 year process to get a properly "Originalist" supreme court, not predicting the end of Roe when there are 5 hard federalist society Republicans is the action of Susan Collins.
Anyway, I would love to hear why you didn't predict the end of Roe in your 2021 or 2022 predictions (I didn't realize you hadn't updated the prediction in the past 4 years) and I apologize for coming off as snide.
I have been reading you for almost a decade and am a huge fan.
How lucky we are that no progressive cause du jour can be cancerous though, right?
Did I say that? Blocking the spread of antivax information is analogous to not yelling fire in a crowded theater. Twitter was right to do that.
It isn't there fault that the majority of famous antivaxxers were conservative.
Of course, there are many terrible conservatives and conservative causes. But since Twitter, and the rest of the Woke Tech, go exclusively after them and are not only lax but often cheer for the progressive ones, their claims to be a force for good or whatever are laughable. They are culture warriors, pure and simple, and the sooner Musk shows them the door the better.
> Blocking the spread of antivax information is analogous to not yelling fire in a crowded theater.
I disagree, and so (I think) do most/all Western countries' legal systems.
Given the general liberal proclivities of Scott's readership, I doubt many of his readers are anti-abortion, even the Republican ones. Most libertarians are pro-choice.
As someone who usually is very critical of rationalists at-large, I agree: there are a few high-profile commentators who are pro-life here (Deiseach, for example, and I think trebuchet as well), but to me they seem to be outliers. I think the "average ACT reader" would like Roe v. Wade repealed on states-rights grounds followed by all 50 states passing legislation that allows for some level of abortion.
Does the "average ACT reader" really believe in "states rights"? I know that many ACT readers believe in rights for many sorts of non-human entities, but states seem like a particularly unnatural fit.
Since you wish to be a pedant:
They tend to prefer smaller, nimbler, more responsive governmental structures on many issues, but aren't outright anarchists. This is commonly referred to in the US as "state's rights" due to the Tenth Amendment being interpreted by them to reserve the bulk of civic administration to the states instead of the federal government.
I would use a somewhat different argument; I don't have strong opinions about exactly what level of government these decisions should be made on, but I do have strong opinions about the rule of law, which means the Constitution should either be followed or changed-then-followed on these matters.
Which, in practice, comes out to being "pro states rights" since the natural tendency of power is to centralise.
On "smaller, nimbler, more responsive" government, I think the US gets it wrong in many cases and devolves powers to an unreasonably low level, e.g. in giving local rather than state governments control over things like police and schools.
As a side counterpoint, local rather than state is most definitely the way to go in terms of control over schools or police. What you would want is a check or balance that is in the hands of the state - standardised assessments that let the local citizens know how their locally controlled school is performing vs others, for instance. We have the pendulum swung the other way in India, i.e lot more centralised/state control, and though admittedly the American govt is better than the Indian I've, the forces that centralisation puts into play are much tougher to handle.
I think it depends on definitions - a number of libertarian people are anti assault to the point of passing laws against it & getting the state involved.
More likely the strong pro abortion stance of the commentariat is a result of the underlying liberal/ left of center make up.
No one is "pro abortion". People are "pro other people can decide for themselves whether to carry a fetus to term".
I'm pro abortion. Those fetuses know what they did :V
More seriously, though, I actually am pro abortion if the child would have a significant disability or if the parent is a teenager or otherwise not financially stable/not in a position to responsibly raise a child.
When in doubt, cut it out.
Except oddly, Ron Paul is anti-choice when it comes to abortion, and he's probably been the most popular Libertarian presidential candidate in my lifetime.
Well it comes down to your conception of human rights. If all humans have rights, then even a libertarian should be pro-life because abortion violates a prenatal human's right to life, and violates it without due process. I mean libertarians aren't anarchists, they do generally believe murder should be illegal. If you believe humans only have rights if they have certain qualities (consciousness, level of physical development, sentience, ability to have preferences, etc) and that prenatal humans lack the relevant qualities then naturally a libertarian would be pro-choice.
So libertarianism is not really that relevant a distincter for pro-life/prochoice, you have to know their conception of human rights and who they apply to.
Here's a poll showing what the DSL portion of the readership thinks: https://www.datasecretslox.com/index.php?topic=6471.msg247766#new
It's a bit hard for me to interpret the answers on that poll. It sort of looks like the answers to that question add up to about 100%, but the answers don't look exclusive. I'm guessing anyone who checked one of those answers is interpreted as meaning that they think abortion should be legal in the case mentioned, but none of the others - but I see 9.5% checked "in case of incest", and I don't think I've *ever* heard anyone endorse that as the *only* case they'd allow abortion (particularly when severe disabilities of various sorts are other options on the list).
>Most libertarians are pro-choice.
I think that's true, but only weakly so and I wouldn't count on it to predict the views of a non-random subset of libertarians. Never mind a group which isn't entirely and maybe not even mostly libertarian to begin with.
Libertarians are pro-most-choices, but they are as anti-killing and especialy anti-babykilling as pretty much the entire rest of the human race - "chosing" to kill a baby, even your own, is a wrongful act that is almost never justified. And whether or not one views an N-week-old fetus as a "baby" with a right to not be killed, is a question lots of people really sincerely disagree about and which is orthogonal to the question of personal liberty vs state power or whatever. Some libertarians do, some don't.
Mmm, it depends. The vast majority of libertarians believe murder-in-the-normal-sense should be illegal, but not all of us derive that from a right to life. There's a categorical-imperative argument that murder-in-the-normal-sense *must* be illegal; people don't want to be murdered, and adults and teenagers are quite capable of arming themselves and/or joining mutual-defence groups (and this isn't even criminal conspiracy if murder is permitted), so a society with legal murder would immediately turn into warlords in an arms race who themselves privately enforce "murder is banned" within their zones of control. Contradiction; QED. So even without actually believing in the right to life, you get "murder should be illegal".
But this argument has a *lot* of exceptional cases where allowing X form of killing doesn't result in the chaos and arms race that allowing general murder does. Duelling doesn't result in chaos, for instance, because someone can just not agree to duels if they don't want to be killed and thus there's no motive to start militias. Allowing kills in self-defence doesn't result in chaos, because someone can just not attack people if they don't want to be killed (with that said, you have to try to avoid cases in which two people can *each* legally kill the other in self-defence, otherwise you're back to at least a degree of arms race). And abortion (or even infanticide) doesn't result in an arms race because fetuses and infants aren't capable of fighting back either individually or collectively. You can to some degree even carve out an exception for killing slaves, so long as slave revolts are infeasible -history contains a few ugly examples of societies that did this and yet managed to endure for centuries, so as terrible as this kind of thing is it's clearly not incompatible with civilisation.
So if you're going absolute hardline "all is permitted that can be while avoiding 'bellum omnium contra omnes'", you get "abortion should be legal". On the other hand, not everyone who calls himself/herself a libertarian goes *that* far (at the very least, most balk at the idea of slavery).
"The future of abortion" sure is a little light on what overruling Roe v. Wade means for people and why we care. Would especially love an analysis on how federal abortion allowance would fit into the Archipelago and Atomic Communitarianism worldview.
re: misinformation, Prediction markets solve the problem of predicting what some authority will say at some future date, but not which authority to use for the settlement source. So it only works if the chosen authority eventually converges on the truth and actually reports it. Big media can't usually get away with lying about simple binary questions of fact, but if there's any ambiguity they can round it to whichever side they feel like and e.g. denounce the lab leak theory prematurely.
(A bit of self-promotion, hope that's okay)
Manifold is now allowing users to send money to charity:
https://manifold.markets/charity
Funny enough, this project got started because of a short conversation I had with one of the founders (Austin) in the comments of a Mantic Monday post!
Yup! This was Sam's idea - so Manifold Dollars are now worth USD to your favorite charity, at a $1 to $1 conversion rate!
Scott Adams predicted specifically that Trump would win in 2016 *in a landslide*. He didn't outperform the prediction markets: his prediction was wrong.
Ditto to this. I'm already skeptical of proclaiming someone a prophet because of a single unlikely prediction (if you make 10,000 wild claims, one of them will probably come true even if the absolute odds for each one is 0.01%), but SA didn't even make a claim that came true! He made a claim that ended up being vaguely truth-adjacent! And I don't think his theory as to why (which boils down to "Donald J. Trump has powers of persuasion and charisma more properly belonging to a comic-book supervillain") reflects reality at all, either. As internal comms around 1/6 have shown, a lot of people in the GOP secretly hated him and only worked with him out of strict political necessity!
Nate Silver's prediction was a *lot* closer to correct than Scott Adams's - Silver was only off by a couple points in a couple states! The problem is that the effect of these differences is very discontinuous, and people are evaluating the correctness of the prediction by closeness in these effects, rather than in the thing itself.
The problem with prediction markets is that the best people won’t focus on events too far in the future. My friends are trading crypto futures. If they can’t make at least 30% pa on good enough volume, they won’t care.
Re. the theoretical Trump manipulation: skewing the prediction market with money might not work, because that can be viewed as an unexpected, unexplained deviation. But manufactured news in coordination with targeted media pieces, bold claims _and_ skewing the market with money might kind of work. Even more so, if a lot of people are watching the prediction markets as a gauge and trusting them, like the modern stock market.
Also, another problem which surely has been raised, but I'll do it again just in case. Insider trading, e.g. Scott buying tons of the "female moderator" option, and then appointing a female. I don't really see the solution to that. It can be outlawed, but will be extremely hard to police. I'm pretty sure this must have been discussed, so if someone could link me that, I'd be very grateful.
Insider trading is often considered to be a good feature of a prediction market - it means that the insider information gets out.
Of course, it's a bad thing when it *affects* the insider decision, rather than just being a way for the insider to cash in on insider information by releasing it through the market.
(For instance, selling shares in ImClone when you hear about the bad trial results is a way to make the price go down, so that more people know about the bad trial results even before the FDA rejects the drug.)
Doesn't the prediction of 10% chance of a female moderator strikes anybody else as just... wrong? I mean, if the user base is 90-10, making that prediction essentially boils down to believing that the next moderator will be chosen at random. That cannot possibly be true... right?
The fact that there are strong female candidates in contention (although we don't know among how many, I'd assume that it is less than 20) means that the probability was necessarily higher than 10%, and the secret better was more accurate than the market -- although I concede that 100% was obviously wrong as well.
Which means that the market going back down to 7% (lower than 10%!) is essentially saying that this prediction market, the one with the highest volume on the platform, is the least accurate (against the commonly held assumption that a higher volume will tend to higher precision). This seems to me the perfect case against prediction markets. Am I missing something obvious?
In every STEM company you've got a lot of men in the bottom layers, the first level of management and the top level. And a lot of women in HR and upper-middle management, i.e. admistrators. I don't think that's just random or sexism. They really like these jobs more.
So for scott's admistrator positions I would expect the density of women to be at least twice that of the surrounding milieu. So a market value of >20%
I give slightly higher than proportion odds to a female head mod for a slightly different reason. Being female protects you (somewhat) from a set of attacks. A bad actor won't blame sexism for their behavior being shut down, you're somewhat less of a natural target if the Eye of Sauron (woke media) decides to cast its gaze in our direction again, and some people will be a bit more cautious in their public attacks lest THEY come off as sexist. It's much the same reason that the WH press secretary in recent years is generally female.
There's a natural counter-argument that being female opens you up to additional sex-based attacks. Certainly the internet has plenty of people willing to toss incredibly offensive attacks at women. But I don't think this comes close to counter-balancing the protection, because those attacks are simply easier to deal with from a tactical perspective. If someone throws out a sexist attack on a female mod, you can ban them with >90% popular approval in this community. If someone accuses a male mod of sexism, you will not have near-unanimous popular approval to simply ban them.
I think 20% might be a better bet-line than 10% here.
It's a belief that the moderator will be chosen at random *with respect to gender.* It's probably going to be nonrandom on other attributes, like "how long have they been a member of the discord?" or "how frequently do they post on ACX?" If something has no bearing on your ability to be a discord admin, then the outcome *should* be random.
What this prediction is basically saying is "I have no special information on which of the candidates would make a better mod, but I believe the people choosing aren't sexist." That doesn't sound too surprising, if the candidates are all pretty well qualified. I'm guessing that "discord admin" isn't really the sort of job where you can be a rockstar who's 10x more effective than the next guy, so which candidate is "best" is going to be effectively random, or based on weird contingent facts like how much they vibe with the selection committee, which is something that you don't really have much information on.
"any more than the stock market is competing with investment banks and market analysts."
I thought they were competing-- that money flows to the more profitable markets. This is complicated by people wanting different things like risk vs. security or getting things wrong, or working on different time scales.
Do we need a theory for why there are different sorts of money markets?
The intended sense of "the stock market" here, I believe, is the *literal* stock market i.e. the physical and legal machinery allowing trade in securities, not the extremely-common metaphorical use to denote "the aggregated people who hold and trade securities".
Elon Musk made a fuss recently about index funds. Saying: if most of the market is index funds instead of people actually trading, then the market price records a lot of noise instead of an actual valuation of the various companies.
There's the same effect for prediction market vs experts. If you have a domain where the correct answer is found in the mouth of experts, then the prediction market is just the opinion of the expert plus noise. And the more people listen do markets instead of directly going to experts, the more noise I'm expecting to see in the markets.
Now I'm imagining a prediction market index fund: they buy a little of every opinion! Which, unlike the stock market, is sure to lose money because a prediction market is zero-sum. So really you should model a prediction market as being a bit like a stock market but with no index funds. No index funds means most traders will either be those with expert knowledge on the particular prediction, professional investors, and casual investors. Professional investors will have hedge funds and the like to be able to throw a lot of money at a market, and will have the resources and incentives to try to bet accurately. Casual investors will produce noise, but generally should be unable to counter professional investors with big pockets. Unless they band together somehow...the GameStop situation from last year does come to mind.
So what’s to stop the outgoing discord admin from betting on his own choice?
Nothing. That's just a way for the market to reveal the answer before the official announcement.
It doesn't strike me as particularly impressive that a prediction market can copy the performance of an expert? I mean, we already have the expert? The market adds nothing in this case.
I think it's likely that prediction markets are overrated, at least until you add insider information, but then that makes for a whole different problem instead.
How do you know which expert to listen too? How do *I* know which expert to listen too?
If everyone already agrees on who has the right idea, then yes, a prediction market doesn't add much. It seems to me that this is frequently not the case!
How does the prediction market know what expert to listen to, in that case?
I mean, *either* the expert is better and the market just copies (and with some delay, even), or the market actually performs better than any expert. In the first case, what's the point of the market in the first place?
Well, if there's anyone good at figuring out which expert to listen to, the prediction market will end up reflecting their views on which experts to listen too...
But the high-level meta-answer is that a prediction market performs boundedly-worse then any process anyone is actually using to formulate their beliefs, without you needing to know in advance to listen to that process and not others, because the best process will get higher weightings over time as people loose money to it.
I think this can sound unimpressive - "what, we only find out who was right answer once everyone else goes bankrupt?", but every trick you can think of to outperform it is also a trick the market can use.
This includes object-level processes like "look around you and try to figure it out", but also meta-level processes like "listen to people with degrees" and meta-meta-level processes like "listen to rationalists about which fields have experts worth listening too and how to find them", and even "infinitely-meta" processes like "notice when you are loosing money on prediction markets and change how you think".
If you, personally, already use the best possible approach to truth finding, you won't get anything out of checking a prediction market. And if everyone already knows the best possible approach to truth finding they won't either. But you probably don't, and they *definitely* don't.
But being better than simply the aggregate of all predictors doesn't sound all too hard? Sure, if an iterated market eventually kicked out all the people with poor judgement (by taking all their money, if nothing else), then it would slowly end up with a competent core that would create fairly strong predictions, but cases like sports betting or even election betting don't seem to bear that outcome out?
Looking at election betting in particular, it doesn't seem to be very good at establishing the *actual* chances of the candidates? We always get a lot of bonkers numbers, and Five Thirty Eight tends to scorn the "Scottish teens".
It doesn't even get you the performance of the best expert; it gives you the consensus of the top experts. Someone might statistically beat it out by an illegible margin.
But that's GREAT for any situation where it's hard or expensive for you to figure out what the actual consensus of experts is on a topic. Take... the Jones Act (which prohibits foreign-flagged vessels from doing interstate coastal trade.) Repealing that should decrease transportation costs (from first principles, increased supply), but unless you've done a deep dive on it you probably don't know whether that's likely to be a 1%, 10%, or 25% reduction in price. A prediction market lets you get a number for the expert consensus on it without spending the cost and effort of doing a deep dive into the field and the people in it yourself.
Is it actually the case that only the experts play? Otherwise, this could easily be interpreted as non-experts muddying the waters with their prejudices.
Sure, it might be irrational to, but it's not like people don't gamble at objectively poor odds.
In a sufficiently large market, there's too much money to be made. People looking to make money will come in and get rich off the amateurs (by correcting the market toward the expert positions.)
Note that this actually happened with the post-election Trump odds. A bunch of people thought there was a >15% chance that Trump would get inaugurated in 2021, even in early January. That market made a lot of people who paid attention to the experts instead of whatever propaganda a lot of money (but was distorted up by the popular amateur opinion.) As markets get larger, the cost to the amateurs to do the distortion goes up, and so the distortion should go down.
The amateurs lost about £40 million on BetFair Exchange on that exact market.
I think that was a case where there really just wasn't enough expert money to overcome the amateurs.
Note: it's not usually the actual experts investing their own money, but "smarts" - gamblers with deep pockets who make their money by betting with the experts against the amateurs.
Yep! Part of that is because prediction markets almost always have bet amount caps; if there are 5 smarts who know about prediction markets and want to make a bundle on correcting Trump inauguration odds, there's a limit to how much they can make (and consequentially how much they can correct the market.) Broader penetration of prediction markets (so there are more smarts), and raising that individual cap should both improve fidelity of the prediction market in cases like this.
BetFair Exchange is a huge gambling market with no caps - they do not have a general liquidity problem. There is a million pounds bet already on the 2024 Presidential election, and that's people locking money up for more than two years. The Australian federal election, in only a couple of weeks, has £2.5MM on the main market, and a few hundred thousand more on individual electorates and some other markets.
They will routinely have a million or ten bet on Premier League football match, and their record single bet is well into seven figure.
There just was so much money coming from Trump supporters after he had already lost that the smarts just couldn't soak it all up in 2020.
It was a liquidity problem, but a really exceptional one. Worth pointing out that the market ended up with well over half a billion pounds bet in total in 2020; this is why there just wasn't enough money to make the market.
Isn't that excessive chance a sign of failure in the market anyway, though? If the idea is that the market should be highly correct, then it *failed* here. That experts (or people trusting the experts) could make money off of it just shows that we should listen to the experts, not to the (clearly unreliable) market? If you trusted the market rather than the experts, you would *lose*.
Right. I don't want to use the boolean "is/isn't a failure" model, but it's certainly true that prediction markets were inaccurately high on Trump's inauguration in Dec 2020. They performed worse than a consensus of experts.
But at no point here have I claimed that prediction markets beat experts. The trouble is that finding who the experts are on a question isn't a free action. It's legitimately HARD for an outsider to figure out who the experts are in a given field, who the cranks are, and consolidate the generally nuanced and broad-scope expert opinions into a specific probabilistic prediction. This can all be done, and if you do it you'll beat prediction markets and can make money on them. (And in making money on them of course, you bring them closer to expert consensus.)
Put another way, prediction markets aren't competing with legitimate experts in the field. They're competing with the Washington Post or Fox News. They're incredibly legible and easy to find summaries of opinion on a topic, biased to more strongly reflect the opinions of those who the 'smart money' trusts (See Richard Gadsden's post immediately above yours here.) The argument in favor of prediction markets isn't that they're a magical truth-finding tool; it's that dedicated gamblers who are making money off the markets are more reliable at finding and presenting accurate predictions than journalists. It really makes sense from an incentives point of view. Journalists lose money when they lose attention. The gamblers lose money when they're wrong.
Indeed, arguably the 'failure' in 2020 was mostly due to a lack of smart money in the system. Almost all prediction markets have features like betting caps which prevent the gamblers from making more than X dollars each on a given market (and so limit the force they can apply.)
> The argument in favor of prediction markets isn't that they're a magical truth-finding tool; it's that dedicated gamblers who are making money off the markets are more reliable at finding and presenting accurate predictions
Right, so ignore the prediction-market odds unless you want to know where the average gambling addict gets their news, but do try to find people with a proven track record of making money off prediction markets and then read their blogs or something.
Wouldn't you have to actually repeal the Jones Act to get an answer that would resolve the market?
Sure. Such a market would probably be set up in a form like "If the Jones Act is repealed before Jan 1, 2023, will interstate transport costs (by X standard metric) drop by at least 10% from their 2021 values to 2023 values?" and resolve inconclusive if the act isn't actually repealed.
Presumably it would only be set up if there was significant interest in getting it repealed, and so a decent chance of resolving conclusively.
But then we wouldn't know until afterwards if the market had good predictions, and once we know, it would no longer be useful. It seems that we would only be able to find the prediction market reliable for similar, repeated events (such as perhaps sports), where a level of consistency and credibility could be established?
The idea is that prediction markets can be validated as a class; if they've worked with reasonable fidelity to expert opinion on questions structurally similar to a repeal of the Jones Act, then we can expect them to have such fidelity in that case as well.
As someone who has been betting on political markets on BetFair Exchange (effectively a prediction market) for nearly 20 years, I'll say a couple of things about market manipulation.
First, someone spending a lot of money on a market to manipulate the price can do so, but they need to keep pumping in money to keep the price up, and the market will attract more and more "smart" money so it gets increasingly expensive to do so. This was done, intentionally, by one of the campaigns for the leader of the Liberal Democrats in 2006, and estimates were that they spent about half a million (sterling), and ran out of money about two weeks before votes were cast. This did, however, probably elevate them as the main challenger, though they still lost - but they would likely have finished third without this.
Second, the weight of each person's opinion depends on how much money they have; the 2020
US Presidential election market was not resolved until the electoral college met and voted (in December). Large numbers of Trump supporters kept the Trump probability around 10% from when the TV networks called the election until after. Like many people, I put in a few thousand to extract an easy 10% ROI, but the Trump supporters were putting in tens of millions and there just wasn't enough "smart" money to push the probabilities any lower.
It is possible that a really large amount of money can manipulate the market, either by a single person or by enough people that have been fooled by misinformation or disinformation. But it is rare and not having a restriction on insider trading (there's no legal restriction) means that the markets are probably more accurate. I know of many cases of political insiders making extremely healthy profits by betting on insider information about likely election results. Campaign staffers often use this to make up for relatively poor salaries.
My issue with prediction markets is that the arguments for them often don't distinguish between theoretical prediction markets with perfect liquidity and an infinite number of participants etc vs the prediction markets that are likely to exist in the real world. You can maybe get close to that with markets on extremely major issues like presidential elections, but that doesn't necessarily transfer to everything else. And proving how it would transfer tends to be elided
What about prediction of Moldova/Romania and Ukraine cooperating and ending Transnistria issue with force?
This is the topic I would really like to know.
Sometimes I think that the best moment to get Transnistria back is now: when Russia is weak, busy, and generally opposed by everyone. After the war in Ukraine ends, if Moldova tries to get Transnistria back, Russia can pay full attention to the issue.
But then I remember that Ukraine is only able to fight Russia successfully because of all the foreign help. Moldova cannot expect to get a comparable amount of help, because the budgets are limited, and Ukraine is a more visible cause. So it could end up with Russia attacking Moldova and taking even more of their territory. For example, Russia could accept the loss against Ukraine, and negotiate something like "we will stop the war now and give you half of the conquered territory back, but you will stop helping Moldova".
Scott, I don't mean to accuse you of dishonesty, just a normal oversight, but the last line of the Discorcasting section seems very close to a "heads I win, tails you lose" situation. Are you saying now, before it's decided, that if a female moderator is selected it would be strong evidence *against* the accuracy of prediction markets? If not, I don't think we can take the low price as evidence for prediction markets' resistance to manipulation.
Question about prediction markets generally: is there a good explanation for how/why they will handle long-term, unlikely events?
For instance, PredictIt offers shares that Bernie Sanders will *not* be the 2024 dem presidential nominee for $0.97. Payout is $1.00, so ~3% increase. Let's say I don't think Bernie has a shot -- I could buy a ton of shares, but I'm effectively locking my money up for *two years* for a 3% return. That's barely better than treasury bond yields over 2 years, and there are probably better places to park my money.
For longer-term predictions with even longer shots, this problem seems even bigger. It seems like prediction markets will necessarily be pretty off for these kinds of predictions, where they cap out at maybe 20:1 odds. Does that sound right? Are there other factors I'm not thinking about that will keep them accurate for this?
It's been discussed in the Mantic Monday posts before: https://astralcodexten.substack.com/p/mantic-monday-1115, https://astralcodexten.substack.com/p/mantic-monday-let-me-google-that
I don't think there's a super convincing answer.
Oh nice, thanks
> Disclaimer that Manifold probably can’t handle probabilities this small correctly and there’s no reason to think 0.2% is more realistic than 2%. It’s not 10% though.
Is there any reason to think 0.2% is more realistic than 0.002% ?
I feel like it's very hard to stick reasonable probabilities on these tails. The question becomes "What's the probability that my mental model of social reality is totally wrong in some fundamental way?" and I think that's very hard to answer sensibly.
I don't understand why you don't cover sports betting, Scott. Sports books are mature prediction markets with lots and lots of money sloshing around in them. Are they good? I dunno, I haven't looked into it, I'm in the camp that all this is kinda dumb. But it seems like a very good arena in which to test your claims for prediction markets. For instance, "public teams" (teams that have chronically inaccurate lines because people like to bet on them independent of their quality; the Yankees, Packers, and whatever team LeBron is on at the moment are public teams) are a notorious facet of sports betting that seem pretty bad for your claims about prediction markets. But I dunno, maybe public teams are a myth, or they were a thing in the 90s that got driven out by the rise of quantitatively sophisticated professional gamblers.
I know you find sports gauche, but if you don't care enough about prediction markets to overcome that, I don't see why I should be asked to care about them either.
Prediction markets kind of remind me of usury. Usury, of course, being the practice of lending money at interest, which was widely regarded in the western world as a sin for over a thousand years. The sin, very generally speaking, was loan sharks: someone loans you money at interest, you can't pay it back, you're in debt to them forever. It was also seen as cheating people: you (the lender) don't actually do any work, so why should you get paid? Today that objection sounds silly: they get paid to compensate the risk of not being paid back, and loans provide a lot of social good so what's wrong with making a profit? But back then all they had as their model of loaning money at interest was loan sharks, so it was condemned as a sinful practice. Today the world economy runs off of loaning money at interest, and the idea of banning it would be almost unthinkable.
So to with prediction markets. Gambling is considered a sin, and for good reason. We all know people who have ruined themselves through gambling, and it's hard to see what social benefit gambling provides anyone besides entertainment. It makes sense that just about every religion and most respectable secular authorities condemn gambling as evil. Yet I can imagine a future, a few hundred years from now, where the idea that people once considered betting on the outcome of events as a terrible sin will be just as surprising as when someone today finds out that lending money at interest was once considered a terrible sin. It's not that our values will change, but the form of the thing itself will change. Just as the free market transforms avarice into societal wealth generation, prediction markets could transform gambling into a prosocial good as well.
> it's hard to see what social benefit gambling provides anyone besides entertainment
Well, this is different with prediction markets. The benefit is having good information, and the possibility to tell people who talk nonsense to either bet on their beliefs or shut up.
Oh I agree. I think that "gambling" now has the same social connotation as "usury" did to a medieval priest. If prediction markets take off then they will do to gambling what fractional reserve banking did to usury.
Gambling a sin? I grew up playing cherry bells and turkey bingo at my (Catholic) church. In high school we had to sell raffle tickets as a fundraiser for the school. Sit down at a poker table in a casino and you'll see a wide range of religious backgrounds represented.
> A lot of times, people ask: what if some very rich person tried to manipulate a prediction market? Like, what if there was a market on whether Trump would win the presidency, and Trump himself dropped $100 million into making it say YES, either out of vanity or in the hopes that people would think he was “inevitable” and stop resisting.
> And the theory has always been: if traders see a clearly mispriced market, they’ll rush to it and correct the mispricing in order to make easy money. Trump would be out $100 million and the market would stay correct.
> We’ve never been able to test the theory, and until someone drops $100 million into a real money market, we never will.
The problem is that you won't necessarily know whether this has been done. It's pretty plausible that Michael Bloomberg did exactly what you said Trump could have done (see https://www.vox.com/future-perfect/2020/2/14/21137882/prediction-markets-bloomberg-sanders-president), but we don't know.
A lot of commenters are making fun of this [someone born before 2001 living to 150] for being too low.
Increases in life expectancy have stalled or reversed recently.
Increases in life expectancy were largely coming not from old people living to older ages, but from young people not dying young, in particular not as infants. It looks like this is something we will reverse - old people are likely going to live to be older. Unfortunately, I do not expect I will benefit from these advances, I was born in the early 80s, but I do think that most kids born these days will live much longer lives than we see today
Well it’s both reducing infant deaths and increasing adult longevity. I think we’ve hit a wall on the latter. We were prolonging sickly old age but even that has reversed. We are not really solving the problems of ageing and some of it looks really bad to me. As in looking at my parents generation (not my actual parents luckily) there’s a lot of people in very bad shape. My age group is looking after them and most of us don’t want to go through that, I think euthanasia will increase in popularity.
Another edition of "please someone explain betting markets to me." In this edition, I made a bet on Manifold and am super confused by the result.
I think there's a solid chance (maybe 20%?) that Republicans will make Abortion illegal nationally within the next 5 years. I saw that the Manifold market was at 6% with $198 bet, so I decided to weigh in. I submitted $25 for a Yes bet.
Two confusing things happened: First, I received 111 shares of "Yes." I would expect 416 shares ($25 divided by $0.06 per share). Why did the shares cost $0.23 instead of $0.06?
Secondly, the Manifold probability jumped from 6% to 78%! That's a huge jump! One, it's above my own true estimate, which is dumb. Two, my $25 bet is about 1/9th of the $223 total market value, so why is my Yes bet outweighing everyone else? Looking at the graph, all the other bets (totaling $198) moved the probability up or down by a couple percent. What's going on?
Guessing: the price of $0.06 was the lowest price anybody was willing to sell for, but there weren't 416 shares of that. So first you cleared the 6-cent shares, then your offer went on to eat up shares that were more and more expensive, until you had $25 of shares. The last of those shares were at $0.78, so that's the new "market price".
So the price is the percentage in cents?
Yes, each individual bet is typically for a dollar, so a 1% bet is 1 cent.
I can see why that market is so erratic.
Did anyone sign up for the forecasting tournament mentioned in the previous open thread? If so, did you get a confirmation email after submitting the signup form?
>Manifold recently ran an unintentional, kind of crazy experiment in manipulating prediction markets.
Alternative possibility, someone knows who the moderator is going to be and wants to have a massive position on that outcome.
This may be a pointless bit of language carping, but I wonder if it's worth considering a slight change in language, away from terms like wrong and right, when you make probabilistic predictions.
"By the way, in 2018, I got this horribly wrong - I said only a 1% chance of Roe v. Wade getting repealed in the next 5 years."
It's not obvious to me that this prediction was wrong! I mean, I understand this use of the term, but strictly speaking, the claim that you made has not been invalidated or disproven by subsequent events. And part of the value of prediction markets is precisely to push people away from simplistic right/wrong pigeonholing of views. One way to do that is to not use the simplistic language when discussing them. Instead you could use something like, "This now looks like a poor prediction; I lost that bet; in hindsight, it looks like I may have evaluated some factors incorrectly." The potential downside is that you sound a bit less human/normal, of course, so it's bit of a trade off.
It's an interesting question whether a probabilistic prediction could ever be wrong. As a dumb example, if you flip a (ordinary) coin and predict a 1% chance that it will land heads, then your prediction is "wrong" in the sense that it's working from an incorrect model of how the world works.
I think it's fair to say that Scott's 2018 prediction was wrong in this sense, that he was working from an incorrect model of how the world works, and I think that most expert SCOTUS-watchers would have been able to tell him that at the time. It wasn't a combination of unlikely events that got us from 2018 to this point, it was a sequence of fairly likely ones.
Thanks, yeah, I take the point about getting the percentages wrong. I'd like to push back on it a bit! I haven't delved into the details yet, but if Scott's 5-year prediction was made at the beginning of 2018, then he would have been right if RvW wasn't repealed before the end of 2022. The ruling is due to come out in June, so his prediction held up for 4 and a half years of the five year period. And there's a lot of happenstance and coincidence that can affect any court case, so the chances of the ruling coming 8 months later are good.
Of course, this may be one of those cases where I'm applying one of Scott's heuristics that work 99% of the time... I'm not sure!
> On the DSL threat, someone brought up that the US government was quicker to call the Russian invasion of Ukraine than most prediction markets. I agree this happened. In some sense, it’s unsurprising; the US government has spy satellites, moles in the Kremlin, and lots of highly-paid analysts. Of course they should do better than everyone else.
I predicted this here:
https://www.reddit.com/r/TheMotte/comments/s2gk0v/will_nato_expansionism_lead_to_a_war_between_the/
I also hinted at Covid ahead of time also in my book:
https://www.amazon.com/Dark-Arts-Rationality-Updated-Digital-ebook/dp/B07ZGHS73F
Given that I seem to have a pretty good track record of prediction, it's a shame that I've been banned from both Astral Codex Ten as well as the Motte. The consistent pattern there seems to be small-minded mods telling me that I'm not NICE enough, but as we learned from the Ukraine invasion, reality is not always NICE. My embrace of reality in preference to NICENESS could have averted a war and a pandemic, but because people don't like taking advice from somebody who isn't NICE, I guess we needed to have a whole bunch of people die instead. Maybe the community should stop optimizing for niceness? It doesn't seem to be a very useful (or rational) strategy.
Please tell how you could have averted a war and a pandemic. Congratulations for predicting it in advance, but how would you have used those predictions to change the outcome, and how did getting banned stop your plans?
I mean, "immediately ban all transit from China, from anyone else with cases, or from anyone who doesn't do the first two" would have fairly-reliably prevented X non-China country from getting CoViD (and in a lot of cases, the public will was there and this was sheer betrayal/incompetence by politicians). And I think he means that "if people had done what I wanted, they would have averted this", which is a whole lot weaker than "if I hadn't been banned from a subReddit, we would have averted this".
"My embrace of reality in preference to NICENESS could have averted a war and a pandemic, but because people don't like taking advice from somebody who isn't NICE, I guess we needed to have a whole bunch of people die instead"
Wow, Mr. Really Big Important Guy, I had no idea you were Emperor of the World! Please tell me why your subordinates in major world governments refused your lawful instructions and why you did not eviscerate them all as punishment?
Or contrariwise, you might be Just Some Guy with an opinion, like millions of others, and why should anyone listen to you claiming you are right, over the other guy claiming he is right that Covid is caused by fluoride in the water?
Oh, you published an e-book? Like hundreds of thousands of other people. Yes, a self-published book by Literally Who has immense credibility.
Make a prediction, right here and right now, that can be checked. Don't do the "I hinted at this in my book" after the fact bollocks, that's astrology. Make a definite prediction that X will happen by time Y, and then we can see if you were right or not.
Get a majority of administrator candidates on board. Buy. Candidates announce they now identify as female. Sell. Candidates announce they now identify as male. Buy...
(Admittedly this works better if the market is real money and you can offer them a percentage of the profits. But I personally would try to get in on the first gender market manipulation scheme just for the anecdote.)
A core issue with prediction markets is that the smartest people in the world are not particularly motivated by money.
If they already have an opinion on the topic, including the probability estimate, it is almost *free* money. Even people who do not care about money too much would probably take free money.
And even if they do not care about free money either, as long as they tell their opinion to someone else, that other person can place the bet for themselves.
The court-packing Metaculus question has moved up since this post was written.
I have not read all the comments / posts / etc related to prediction markets, so apologies if this is redundant/covered. But I have not seen it covered in what I have read.
Prediction markets seem neat now, because they seem to be high-signal. People mostly seem to want to predict, not manipulate. People mostly want to be accurate according to factual reality.
This seems to mirror the early Internet. Very less overall content but higher average quality. But then came mass adoption, Facebook, Twitter. More recently, concerted, sustained disinformation.
I would be concerned what happens if prediction markets become less niche, more noticed: "successful" by some metrics. Say Russia or the CCP or some other government-level entity with enormous resources takes an interest in manipulation. The goal here might not be to drive the outcome, merely to muddy the waters and maximize noise. So you could not simply get rich by disagreeing.
on Xi re-elected CCP General Secretary: as somebody who bought No on Polymarket, I see the main possibility to be him retaining power but with a different title. Think of Putin's 'castling' with Medvedev as a comparison. It's not probable, but IMO it's over 10%
Why do you think markets didn't even react to the news of Putin's health? I personally think that's the biggest predictor he might stop being president
https://polymarket.com/market/will-vladimir-putin-remain-president-of-russia-through-2022
Re aging:
Aubrey de Grey says that there are seven (and only seven) things that make us age, and he has solutions (for some understanding of "solution") for all seven.
So
(a) do you trust Aubrey de Grey? Do you trust him more or less after seeing a picture of him? Opinions, as they say, vary.
(b) do you trust that his solutions will actually work (I'm assuming the next hundred years or so are long enough to iron out any engineering issues; the conceptual issues are the point)
(c) [and this is the big one] do you believe that "society" will allow these solutions to be tested then implemented? I flip back and forth on this. On the one hand, rich people and other countries are going to be happy to let this go where it goes. On the other hand, "society" has a long history of being fairly successful in slowing down or suppressing certain types of medical research, especially stuff that requires large scale effort not just a mad scientist in lab (eg much interventionist genetic medicine). So...