A couple of weeks ago I ran a betting market predictions game night for the NYC rationalist weekly meetup (all bets should be settleable by the end of the night and the winner is whoever has the most net points). It is (a) a lot of fun and (b) incredibly chaotic, would strongly recommend more people do it. I wrote a post about how I set it up technically and what I'd do to run it smoother here https://shakeddown.wordpress.com/2022/08/04/what-i-learned-about-running-a-betting-market-game-night-contest/
You get some incredibly weird phenomena - I had one question that was "which question will have the most bet on it", which led to two other questions getting a ton of mana sunk into them as the people who'd bet on them tried to make their questions win out. One of these questions was "which answer will get the most points" (the other was "how many light years away is alpha centauri"), which was won by one guy suddenly sinking all his points into the runner-up just before it closed (and since it had a lot of points invested, he ended up winning the whole game by like a 2-1 margin over the runner-up thanks to this. This was also the guy whose day job is a Jane Street trader).
(Also several other fun attempted manipulations. It really is a great creativity generator).
The 55% chance of war with China by 2050 is honestly surprising. I mean, I perceive China as an extremely rational entity ( that is for a government, governments overall are far less rational than people) and I would assume that the disadvantages of waging a war, should be obvious enough that no rational agent would try it.
But then again, this was the assumption with Russia, that Putin would be rational, see that he does not benefit from starting a war, and not do it. And here we are now.
If you model the regime's utility function as not "make china maximally prosperous" but instead just "keep the regime in power", extreme covid policies could be a sort of flypaper strategy to locate and eliminate potential dissidents before they present a real threat to the regime.
I'm sure some Chinese person is going to pop by and tell us this was exactly the strategy employed by some emperor, perhaps even known by a clever animal name. They've got a few thousand years of history to draw on, and there isn't as much new under the sun as we like to think.
The more I read about China, the less it looks like a stereotypical dictatorship. They have transitions of power, successfully, as scheduled. The old guys are happily retired. The selection process for top government positions is arguably more sane than in western democracies, in the sense that it's not 90% a popularity contest and requires actual experience and competence for the job.
Is keeping power a major consideration for the leadership? Well, I'm pretty sure that's true for a supermajority of political systems everywhere - and in democracies it's touted as a feature, not a bug.
The question is if they are "cheating", aka doing things that meta-influence the normal political process. And, at least at a naive look, neither Covid policies nor long term geostrategic goals do that. For a comparative look, mail-in ballots in the last US elections were a significant meta-game move, as they changed the electoral process shortly before elections. AFAIK, China doesn't have anything similar to that - every move they made (good or bad) is only affecting the political process at object level.
What appears rational to Western observers is often countries that do not insist on exporting their particular ideologies to the rest of the world, as the West does.
Xi and Putin have little interest in the internal affairs of other nations, unlike the democracy and human rights-supporting West.
But don’t mistake that lack of principle for rationality. China and Russia put an enormous amount of stock in not being embarassed on the global stage. Both are also dependent on rabidly nationalistic and misinformed publics, and maintain mind-numbing propaganda to that end. Unfortunately for them, the nationalism they invoke also limits their governments’ ability to make unpopular decisions, in some cases more than democracy would. Lastly, Russia and China are kleptocratic states that rely on the support of a variety of elites, either oligarchs or party members. These people are individually rational, but in practice this means they are very corrupt and rely on the state for their wealth.
So no, China is not particularly rational. In fact, I would argue that they very valuelessness that makes people believe Russia and China are somehow more rational than the West is actually a weakness, since they are only able to build alliances of convenience, rather than alliances of shared values. Witness China’s abandonment of Russia compared to Western support for Ukraine--hardly a close Western ally. Pure self-interest is an untrustworthy trait, and tainting that purity with the pride of nationalism hardly makes a state more trustworthy.
> Xi and Putin have little interest in the internal affairs of other nations
Putin appears to be interested in internal affairs of Ukraine, Belarus, Syria, and a few others. Did you mean that he is not interested in internal affairs of nations far away from Russia?
Not really, no. I mean quite exactly he isn’t interested in the internal affairs of other nations. By “internal affairs,” I mean the type of government, the laws, the degree of respect for human rights, the official ideology, and so on. In geopolitical speak, this is a shorthand for a variety of assumptions that come with realism. By “interest,” I mean caring about those facts in themselves. I do not necessarily care about your life merely because I interfere in it for my own self-interest. Psychopaths manipulate and interfere with people, but to say that they care about people is simply wrong. Realism argues that states act psychopathic, always choosing whatever makes them most secure.
Putin’s interest in Belarus and Ukraine is in making them Russian, or Russian statelets. Xi’s interest in Taiwan is in making it a part of the PRC. Neither leader views these nations as separate from their own, regardless of international boundaries. Even if they did view them as separate, their main reason for conquest is not actually some sort of universal principle, but nationalist-fueled revanchism. Russia wants Belarus and Ukraine to be Russia, regardless of the internal politics of those nations.
You mention a potential caveat of:
>Do you mean that he is not interested in the internal affairs of nations far from Russia?
No. Distance is not the relevant factor here. Take Kazakhstan and Pakistan as alternative examples of nations close to Russia and China respectively. Russia cares about how Kazakhstan is governed only insofar as ethnic Russians there are protected and Kazakhstan does not oppose or threaten Russia. Whether Kazakhstan is a democracy or not is entirely irrelevant to them. Whether Kazakhstan follows the teachings of Alexander Dugin is irrelevant to Russia. Pakistan and China are much the same. China could not care less about whether Pakistan is run by a military junta, democratic Islamists, secular elites, or a communist party. What China cares about are the strategic and economic implications for China. These nations are close to China and Russia, but neither country cares about the internal affairs of the other. What matters is solely how what happens in each country affects China or Russia.
Mongolia is perhaps the best example of this. It is a healthy democracy in the midst of the two nastiest regimes of the 21st century, and yet is on good terms with China and Russia. The internal affairs of Mongolia are irrelevant to both nations because its external relations with each are already excellent.
As for Syria, Putin has zero interest in the internal affairs of Syria. It may look that way, but what matters to him is that the regime in charge is friendly with Russia, as the democratic system has assured in Mongolia. The only Russian attachment to the Assad dictatorship itself is instrumental. It is not about how Syria is governed, it is about how Syria relates to Russia. This is an external, geostrategic consideration. So long as whomever governs Syria allows Russia to use it as an airstrip, Russia cares not a whit what happens to the place.
The point is that neither Russia nor China interfere in the internal affairs of other states for ideological reasons, except when they think that those states are actually secessionist members of their own nation. They are not interested in those internal affairs that are truly internal, and lack any geostrategic consequences. Their elites do not worry about the advance of “Xi Jinping Thought” or “making the world safe for Duginism” as the West worries about human rights and democracy.
The best counterexample of this are Israel and Saudi Arabia, excellent Western allies with sub-par and extremely deficient human rights records respectively. Strategically, there is no reason for the West to care about this. But in fact, constant pressure campaigns are applied, the threat of sanctions looms constantly, and international condemnation is common.
Nations like South Africa are forced by Western nations to change their internal affairs (end Apartheid) despite their loyalty to the West if the West finds those affairs too offensive. I think this is obviously good—apartheid was wrong, and it was right to use sanctions to force its end—but it hardly benefitted the West. Indeed, it turned an anti-communist ally into a neutral party (South Africa joined the Non-Aligned Movement in 1994). The Soviet Union once cared about promoting the global struggle for communism, but neither Russia nor China are such ideological projects.
My apologies for the lengthy reply, but it seemed necessary to provide definitions, context, and examples, particularly as you seemed somewhat unfamiliar with common foreign policy jargon.
The problem isn't with China starting war but with West not being interested in preventing such war. By asserting Taiwan independence West makes moves that limit alternatives (like Taiwan being peacefully re-integrated into China - which is still main Chinese plan) and thus potentially making war the only remaining option that maintains China and Taiwan as one country, even with all the costs attached to it.
Cost of war are high but not to the point of "no rational agent would try it" - and it can be choice between "being slowly strangled through other means" vs "short but powerful struggle which will shift battle to more favourable terms even if it is not guaranteed to provide outright victory".
>The problem isn't with China starting war but with West not being interested in preventing such war. By asserting Taiwan independence West makes moves that limit alternatives (like Taiwan being peacefully re-integrated into China - which is still main Chinese plan) and thus potentially making war the only remaining option<
Yeah, that's how I look at the world too: if you don't give in to my demands, I'm forced to fight you... so *you're* the problem!
"Maintains China and Taiwan as one country"? Are you a Chinese propagandist? China and Taiwan are, very clearly, not presently one country. China does not want to maintain the status quo. China would, instead, like to take over Taiwan's wealth and economy and eliminate the freedoms of their citizens and their government institutions.
They are considered to be one country by both countries - that's the status quo. "One China policy" was echo of that reality.
They only disagree on who should have an upper hand in relationship.
Obviously mainland China is stronger economically right now, and thus absent other influences will eventually prevail; West tries to put their hand on the scale and make Taiwanese democracy win due to Western ideological leanings.
This force has to be met with counterforce to preserve previous trajectory - and China certainly seems prepared to employ it.
The Kuomintang (Chiang Kai-shek's party, and the only permitted party in Taiwan until the 90s saw actual democracy) supported one-China (i.e. still claims mainland China as its territory currently occupied by rebels). The Democratic Progressive Party, which is Taiwan's current ruling party, supports Taiwanese independence.
They've so far not had the balls to actually formally declare independence, because the US has told them not to and the PRC has officially and specifically said that it would invade in case of a Taiwanese independence declaration - pissing off the USA while crossing a PRC red line is not pragmatically very wise.
But in the event of an invasion actually occurring anyway, the US would stop discouraging them and the DPP would declare Taiwanese independence.
But until they actually declare independence invasion is highly unlikely and China can continue believing they'll get them eventually in a peaceful way.
So anything that increases chances of Taiwan declaring independence - be that support for Taiwanese independence idea through (previously unheard of) high-level state visits or military support and training similar to Ukraine to increase costs of invasion - increases potential for war.
>But until they actually declare independence invasion is highly unlikely and China can continue believing they'll get them eventually in a peaceful way.
No. The PRC torpedoed that when it revoked Hong Kong's autonomy. The Taiwanese do not want to live under the Party's boot and cannot trust the Party to keep its promises. The only remaining possibilities for reunification are 1) the PRC liberalises and the Party is destroyed, 2) the PRC invades.
The PRC has, indeed, believed for many years that "we'll do it eventually" on a lot of fronts. But if you haven't noticed, Xi seems to have decided that "eventually" is coming around at last.
I'm not going to accuse you of being an enemy agent like REF did; Paul Keating made a similar argument recently and his credentials of "not being a CPC agent" are pretty impeccable. But it's outdated. Yes, they bided their time. Past tense. Now they intend to collect.
It's not quite as simple as "the PRC wakes up one day and decides to start Armageddon".
The scenario looks something like this.
The PRC line has been that "Taiwan is ours and to hell with what the Taiwanese think, we must invade if peaceful reunification isn't happening". The main thing stopping that is US threats of intervention. But the read I get on PRC thinking is that they really do not get the Western moralistic mindset, so I'd expect their guess at the chance of intervention to be significantly lower than what it actually is (and there's always the fascist overconfidence to consider). So there should be a lot of probability mass on "they attempt an invasion and the West does intervene".
Yes. Honestly, this is just a very real application of what Scott has called "the principle of charity," and if you look for it, you can see it everywhere: not taking moralistic claims seriously. The more cynical you are already, the more prone you are to project that everyone everywhere operates on cynicism. But even not-so-cynical people are vulnerable to this bias if they're sufficiently partisan when thinking about their enemies.
The two most likely triggers for an invasion I can see are this: Taiwan does something that makes the PRC feel forced to act and roll the dice (i.e., formal declaration of independence). Or the US is seemingly paralyzed by a constitutional crisis (most likely centered around an election) and PRC feels a narrow window is opened for action.
The latter could work, but it also seems ripe for a miscalculation in which the US ability to act is underestimated. Kerensky's Russian Republic was somehow able to more or less continue fighting WW1 (albeit ineptly) for several months despite the fact that Russia for all intents and purposes did not have a functioning government at the time. Something tells me the US military + intelligence machinery could still act against China for a period of time despite the US Federal Government being apparently paralyzed, and by the time the US military was running out of steam and needed to requisition more funds, the crisis might well be resolved and the government functioning again.
The US threat of intervention might no longer be the deciding factor in hypothetical invasion scenarios. Air supremacy and naval supremacy had a good historical run, but current warfare (including the one in Ukraine) is heavily reliant on missile supremacy.
Taiwan is (even on its own) lavishly equipped with missiles that can stop China's naval advances, just as China is lavishly equipped with missiles that can stop US Navy advances. Any attempt at a direct invasion of Taiwan would be an unmitigated bloodbath, with or without the US.
I'm curious about what you're proposing. A blockade? Or perhaps just laying waste to the place?
The thought I've had is that if the PRC is largely interested in Taiwan for economic reasons, it doesn't require much Taiwanese resistance at all to turn the invasion into a pyrrhic victory. It requires Taiwan to surrender rapidly and China to be very surgical in its bombardment.
But if China's interests are more a matter of prestige and making an example of the place -- costs be damned -- then it can safely lay waste to the civilian infrastructure and starve the island into submission, unless the US is prepared to break the blockade (which was touched on in that CNN article). In this scenario, China might not even be interested in conquering Taiwan -- simply bullying it into vassalage would be sufficient. Let Taiwan pick up most of the bill for rebuilding its own infrastructure.
When it comes to China attempting a sudden, swift conquest, the most important contribution of the US might be to morale. There has been a lot of doubt cast on the reliability of Taiwan's military and reserves, though I think everyone has revised their appraisals upwards after witnessing Ukrainian resistance. Still, the perception of whether or not the cavalry is coming could prove decisive to the marginally motivated ROC soldier.
The PRC is not largely interested in Taiwan for economic reasons. They were willing to wage bloody war over Taiwan when Taiwan was an economic basket case, and the reasons for that have not changed.
The CNN article misses something re: blockade. Sure, the USA wouldn't go in shooting just because of a blockade - but they wouldn't have to. They'd just *run* the blockade, and throw the dare back in the PLA's teeth with "are you going to shoot at USN ships in waters not recognised as yours". Similar to the "freedom of navigation" missions they've been doing for years - they sail through what the US says are international waters and the PRC claims as territorial waters, and basically dare the PLA/PLAN to do anything about it. Or, for that matter, similar to the Berlin Airlift.
And yeah, obviously the PLARF could nuke the island into oblivion if taking the industry intact was no concern and the West responding in kind was no concern.
"Running" the blockade /does/ mean shooting at ships, not making rude gestures at them as they pass by. The blockade *is* military warfare, not economic one.
As that CNN's comment says, "What's cheap is a surface-to-ship missile, what's expensive is a ship." PRC can easily do "surface-to-ship" to any Taiwanese ship (or plane), and so can Taiwan to any PRC ships in its vicinity. Taiwan can't do much surface-to-ship stuff to PRC ships departing from ports adjacent to Vietnam or Korea. And it's clear that surface-to-city stuff done by either side would be immediately reciprocated by the other. Every military war has escalation levels, and that escalation wouldn't be taken lightly.
The West shooting at PRC ships /would/ mean engaging in military warfare, of course. I don't see that happening; it would remain on the level of economic and media warfare.
A military-enforced blockade of Taiwan is a much more sane option for PRC than a military invasion. By "threat of Western intervention" do you mean that the West would actually declare a (military) war on PRC if PRC blockaded Taiwan?
Nah. Simply no way they would, not over a blockade. But they certainly would escalate the *economic* war. (whereas the West's *media* war on PRC and Russia had already been amped up to 11 for at least the last dozen years, there's not much left to escalate on that front anyway)
I think the threat of this economic war *is* what's effectively keeping PRC at bay, for now. The time is not right. The US dollar has had a virtually monopolistic stranglehold on world trade in the post-WWII era. But it seems this is slowly beginning to change. The US sanctions on Russia have caused the first international oil trades to be settled in yuans instead of in dollars - a watershed moment in that area. Biden's money printing is also not helping the dollar's popularity abroad. When those cracks have expanded sufficiently, then PRC might decide that the time is right to endure an economic war with the West.
China has had a fairly rational approach to foreign policy in the post Deng era but now with oligarchy being replaced by dictatorship we should expect less of that as the whims of particular leaders can have more play.
Deng wanted Superpower China too. He just said "bide your time until the PRC is strong enough, because it isn't yet". Now, 30 years later, well, it's quite a bit stronger and so they're thinking about getting on with it.
Unlike Putin's invasion of Ukraine, which was unambiguously illegal under international law, Taiwan is legally part of the PRC and therefore they have the right to invade it whenever they want, just like the USA has the right to stick as many troops in Hawaii as it feels like.
Also, you are looking at the calculus for China invading Taiwan today. Right now, the USA unambiguously has a stronger navy and its sanctions could easily be crippling, but China is developing quite rapidly economically. Say it's 2040, China and the USA both have domestic semiconductor industries as good as Taiwan's, China has an economy the size of the EU and US put together, and China's navy can compete with the USA in the Pacific. Is war still a drastic miscalculation?
Sure, that would also be fine under international law, in the same way we've adjudicated countless civil wars. However, unlike the PRC invading Taiwan, it has about a 0% chance of happening.
This is apparently a reference to the Sword of Chang, which is conjectured to be a confused reference to the idea of Chiang Kai-Shek reclaiming mainland China. Scott mentioned it (with some links to articles) here:
While the law is clearly much less effective in international relations than power, I think it is also too cynical to suggest it means literally nothing. There is a reason why the PRC obsessively polices people's adherence to the One China principle.
China's government uses CCP ideology as a co-ordinating glue that holds everything together and reduces the odds of late Soviet terminal ennui. That co-ordinating glue has as one of its central planks "reclaim Taiwan by 2049".
This clock was set ticking by Mao talking to Nixon:
More dramatically, there's a story (whose link I can't find right now) about a high-ranking US official from the Clinton (or Obama?) administration having a chat with with a high-ranking CCP official at the sidelines of some conference. The US official casually said "And Taiwan is still a thing we can punt for 100 years right?" and the CCP official replied "No, that was X years ago, so now it's only 100-X years"
At the time, America interpreted 100 years as "forever away" and the CCP interpreted 100 as "actually 100". Given that the CCP used the terms of a 99 year treaty to reclaim Hong Kong in 1997, this is relevant.
I think calling "more than 100 causalities" a war is stretching it. I would call it a skirmish. You could easily get to that number by some naval ship crossing into territorial waters and then being sunk.
Even a conventional war would probably quickly rise in the tens of thousands of death, at least.
Given that when one hovers over the picture, the pop-up text includes a title that identifies the portrayed figure as "Kalki", it seems likely that Scott is aware of this.
Which is why I chuckled at what seemed to me to be a joke based on Scott purposefully making fun of the American tendency to confuse foreign words, while engaging in a little bit of hyperbole associating Kalshi with apocalyptic things.
Re "I’m kind of confused what’s going on here", I would guess that while there's a lot invested, there's not many limit orders on the 'no invasion' side (because China might invade Taiwan any day, or there might be news that it's suddenly much more likely to do so, and then the probability spikes for real so your limit order is filled but it's now a bad investment). You don't want to leave open an offer on 'no invasion', but you'd happily buy 'no' if you see that the price is way too high. Thus one buy on 'yes' can change the probability a lot.
Does anyone else enjoy predicting what would happen in fictional worlds is the laws of narrative and plot were suddenly suspended? Like Palpatine says "execute order 66" and then suddenly Star Wars stops being a narrative universe and just proceeds in a way that is dictated by its own internal laws, but no tendency towards narrative arc- what happens?
The difficulty is that in some fantasy universes, the narrativity is written into the plot. In Star Wars, very explicitly and in Dune, implicitly, what happens is the outworking of a superhuman intellectual agency.
The most fun to be had as a dungeon master, I think, is to make sure the player characters have concrete motivations, give the NPCs sometimes-aligning, sometimes-opposing motivations, set a bunch of internal laws about what strategies the NPCs will follow, and just simulate the world as accurately as possible, without regard for plot or narrative arc. Doing it that way actually produces a far more interesting, satisfying and unexpected narrative arc than what I would have written, in my experience.
I think this is because the human mind has an amazing capacity to find meaningful patterns in events, and even when the events are randomly-generated or accidental, those patterns will appear on their own and start exerting a gravitational pull on everything.
I once (at age 21?) ruined (or made *amazing*) a 12? hour DnD session. We were captured less than an hour into it, and I spent the entire time trying to escape a jail cell when the DM stated some logical inconsistencies and refused to back down from them. Many laughs were had.
Specifically we were served food through a slot in our cell door, but no one was emptying out toilet buckets. But the bucket were emptied...
So there were hours pushing the DM about how the magic shit buckets worked.
In 2016 on PredictIt, I saw one case, which seems like undeniable insider trading. I was on the RealClearPolitics poll market, and someone was buying large amounts of a 2c YES stock, 10 minutes minutes before closing (around midnight EST). A few minutes later, a new poll was added to RealClearPolitics, causing the correct answer to flip. Even with the $850 market cap, buying in at 2-5c, would mean $20k+ in winnings.
In prediction markets is insider trading a bad thing? I thought the whole point was to produce accurate predictions by incentivizing people to use any accurate knowledge they had. Sounds like your insider trader was moving the market in the correct direction. Or was the 538 post inaccurate?
It's a hard and complicated question. My intuition (with no policy implications) is we want markets to incentivize doing research in advance. An insider finding out the answer two seconds before the market closes when info gets publicly released provides almost no value and invalidates the expected return of all those who put in research adding liquidity to the market in the first place.
> [A]s far as I can tell this is only about who makes it to the finalist round, and then to some degree the usual criteria will start creeping in again.
Yeah, that's what it says.
I've been doing some of this prediction market stuff and have been pretty good at it. (I only predict things I have domain knowledge about. But I'm told that's not an issue.) And this sort of thing keeps me out. It keeps me out of a fair number of things where I know there's going to be the same old standard filter applied after all the fancy new ways of sorting. Layering an additional test on top of the standard one only selects for people already in the circle. So if I'm not in I expect I can beat them all and still lose on points to the Harvard PhD.
Another way to think of it: Imagine a normal fellowship takes 100 applications and selects 1 scholar based on normal academic criteria. They disqualify 60 of them and I'm in the 60 because I don't have a PhD. But they chose one of the 40. I won't apply. Imagine this fellowship takes 100 applications and uses prediction markets to narrow it down to 5 scholars. And then selects based on normal academic criteria. Even if I'm 100% confident I can get into the top five, even I'm 100% confident I will be number 1, my expectation is that I get into the top five and then don't get the scholarship. So my expectation is that both of these contests end with basically the same result. The only thing that shifts is possibly which of the 40 who would make it through the original filter would win. The only way it won't is if all 5 winners are not people otherwise qualified which is unlikely. That's something but not enough for me to feel like I'm actually in the running.
The net filter is at least somewhat equivalent to asking people for resumes, selecting the qualified ones and excluding everyone else, and then having them do a prediction market to decide who wins.
This is without getting into the tournament criticisms here which I broadly agree with:
Do you think this provides a way to pad your conventional resume with your unconventional prediction market skills? ("Your" is meant to be generic, I have no idea if you personally are interested in that sort of thing.) Superficially "I was officially recognized as one of the finalists for this cool fellowship" sounds like a resume-worthy accomplishment -- but maybe anyone who'd care would already know that the fellowship in question is kind of weird and therefore not care.
I do not. Firstly, I expect most people to not know what this is and in the worst case to see it as me bragging about being a good gambler. And secondly I expect most other rat-adjacent jobs to have the exact same issue as this contest. They're not searching for the best predictor. They're searching for the best predictor who's a graduate student of an elite college or similarly impressive.
In this the rationalist movement seems significantly worse at finding underappreciated talent than several large corporations let alone some more insurgent movements. Instead its primary ability is to attract a certain kind of high class talent who'd otherwise work at Deloitte. Which isn't nothing but certainly isn't so much rationally getting past credentialism as running a hype train.
Isn't "bragging" what resumes are for? (You know much more than I about rat-adjacent hiring practices -- I'm both credentialed and not rat-adjacent -- so it's interesting to learn about.)
I mentioned this. It's not impossible but it's unlikely. And even if I get through the question is then whether I'm the best of the five and so relies on my estimation of the median credentials of who's involved. Which I expect to be generally better than mine.
Same thought here: why bother? This is hardly any different than a regular application process, at least insofar as "only apply if you think your resume is good enough to get the job" — how you do on the prediction competition isn't really a factor unless you *already* believe you're a likely hire.
Slightly OT, but a more successful job search strategy (outside academia) for me has been to apply if I think the chance of getting *an interview* is worth the effort required (e.g., "easy apply" roles are less effort, so fire away). Your chance of getting the job, conditional on getting an interview, is largely independent of your resume.
Yeah I was disappointed to see the format of the tournament, likely no more predicative than a March Madness Pool, where that one random fan who loves terrible team X wins for whichever team X surprises.
If the MATH dataset is open source, is there no guarantee people aren't just overfitting the test set to meaninglessly pad their accuracy in ways that won't generalize out of sample? To run a real ML competition you would need a secret test set that the participants never get access to.
So I'm tentatively suspecting that overfitting is the more likely explanation than AI capabilities proceeding 7SD faster than the forecasters expected. It is trivial to just memorize the answers to the test set.
ML benchmarks are often run with public test sets, so yeah it's sorta in the honor system. On the other hand, it's been demonstrated that while the community has overfit e.g. CIFAR10's test set in that the accuracy will drop if you resample the data, higher performance on the CIFAR10 test set still predicts higher performance on imagenet (of the same architecture, not the exact same weights).
If you think about it, you get such slow feedback from the test set and it's so few bits that it's really hard to imagine actually having your model implicitly memorize examples through hyperparameter settings.
People reading this might come away with the impression that PredictIt was still mostly in compliance with the no action letter, since Scott chose to only highlight three of the nine requirements in the no action letter. Just to be clear, and as I suspect Scott is aware, PredictIt was in substantial violation of a number of the other requirements in the no action letter, such as the requirement to keep the subject matter of the markets to a narrow range of subjects and limits on overall revenue generation.
Which markets were outside the subject range? I see only politics and a couple of economics markets on PredictIt. The original letter listed said it would "include" three specific things in politics, but given that PredictIt has been doing more than that forever I assumed there was an understanding that they could branch beyond that as long as they stayed within politics.
What was the limit on overall revenue generation? I mentioned the issue with Aristotle being for-profit, I don't see any other limits (eg specific dollar amounts) mentioned in the letter.
I am not really convinced they're violating the letter, especially because if they wanted to violate the letter there would be much more profitable ways for them to do so (eg removing money limits, removing trader limits) plus they made some changes (like closing their popular tweet markets) that I interpret as regulators telling them to do things and them listening. Probably this involved some limit-testing and limit-pushing, with them taking CFTC not caring as encouragement. My model is that they have tried to comply, have been modulating their exact compliance level based on interactions with CFTC over the past decade, thought they were in compliance with the law-as-it-really-exists-and-is-enforced, and this came as a surprise to them.
PredictIt was certainly aggressive with the $850 limit - they'd routinely list 2+ virtually identical contracts each with independent $850 limits. But as you say they've been doing that forever.
I am surprised about the question regarding pop songs being composed by AI; that is, I was assuming that they were already mostly composed by AI. Was I wrong ? Are there really a bunch of human composers still toiling in the pop mines -- or are humans just sitting at the top of the loop, clicking "next step" when they find a generated song that they like ?
Most pop songs are definitely made by people at the moment lol.
It is complicated, the person’s whose name is on the writing credit may have bought samples from several people and combined them into one tune, but music is made by actual humans.
AI music is currently rather poor, especially if it includes human vocals or needs to have human vocals put over it.
My reaction was "they expect there to still be a Billboard Top 100 by 2050?" Music is already splintered into a thousand sub-subgenres; by 2050 Son-of-Spotify will be producing personalised DJ mixes for everyone based on their mood and biosigns, or even totally novel personalised music. But I guess institutions linger, and there will probably still *be* a Billboard Top 100 even if hardly anyone listens to the songs on it.
I think some people like the fact that they listen to the same music other people do. The specific mechanism in the future may be quite different from now... maybe all music will start as a personalized production, but then people will share it on social networks, and you can have a list of "songs that received most likes"... but it still may be some form of Top list.
Billboard has been integrating streaming music stats in its charts for a long time now. And while there are subgenres, there still doubtlessly is the mainstream, particularly relevant to people who aren't music enthusuiasts, with widely known smash-hits still emerging regularly.
True, but a kid born today will be 28 in 2050. The generation that will be defining musical tastes in the 2050s will have known streaming and algorithmic recommendations all their lives, and think of them as normal. They'll probably have a very different relationship to music than today's tastemakers.
Plausible, but I would expect there to still be tastemakers of some sort. Algorithmic recommendations aren't arbitrary, to be effective they have to reflect underlying human nature, with its herd behavior tendencies.
I always thought I was alone in thinking "basically all of the pop music after around the year 2000 sounds like it could have come out in the year 2000". Turns out it's at least a fringe or minority opinion.
Maybe pop music simply achieved its long sought holy grail of lowest common denominator around that time, and there's no more "improvement" to be had? An insightful video along these lines:
I think a challenge for AI-created music is that there is already an oversupply of music, because lifetimes' worth of music have already been recorded and can be listened to instantly, and a large number of extremely talented people are working for pennies to record music faster than it can be listened to. Next time you're in Nashville, ask your waiter if he's on Spotify.
My first thought is that AI-generated music, to be useful, would need to be applied to very-specific applications that no actual musician will touch, at least not for a price that works. For example, suppose I'm trying to study for an exam, and there was software that could transform information I need to memorize into a catchy jingle? Or my factory wants to do the same thing regarding its safety procedures?
1. I don't think they're actually making that much profit on fees.
2. They are trying to do this.
3. As currently designed most of their contracts are on US politics; I think "foreigners who want to bet on the US government and aren't already using other foreign prediction markets" is a pretty small market.
But they still have six months until they're shut down, so maybe they will do one of these things.
"This seems underpriced, unless people expect whatever group manages the list to disqualify AI songs, or some kind of civilization-wide agreement that they don’t count. There’s already mediocre AI-produced music, and nothing stopping it from getting better with really simple scaling. Given that most songs are already composed by nameless corporate composers and not the pop stars themselves, I don’t think there will be any resistance to this."
Uh....
Pop music isn't what's best, it's what's pushed. It has to be good, but more importantly it has to be pushed, and it has to have the kind of cachet to be pushed.
So this is essentially a bet that the cool crowd of cool kids will decide it's cool to push computers as artists instead of exclusively pushing other (relatively talented and convincing and possibly interestingly insane) cool kids. Will "this was made by a computer, oooooh" be competitive with staring into the absurdly compelling eyes of the likes of Taylor Swift and Billie Eilish and Adam Levine?
You know I'm down on AI, but 43% seems perfectly priced to me.
This isn't about performers, though, it's about composers. The current #1, _Break My Soul_ is promoted as "by Beyoncé", but she was one of four songwriters who collaborated on the song, two of whom I've never heard of. Same deal with #2 - you have to go down to #4 before you find a song written solely by the featured artist (Kate Bush), and the next one is at #9 (Nicky Youre): https://www.billboard.com/charts/hot-100/ So I think it's likely that we'll see AI *composed* music long before we see Idoru-style *performers* - the audience will neither know nor care that North West's latest banger was written by an AI.
Random, but since you pulled me back to 80s music and your name is Chaz, my brain really tried to convince me that you were Chaz Jankel, writer of "Number One" and other 80s fare.
That's true! If the condition of the bets allow partial or full credit for writers as distinct from top of line performers, then yeah, this could be undervalued.
Hatsune Miku is evidence that you can push a computer as a pop artist, although getting 50% of the Billboard top charts is a very high bar to clear.
(Actually, Miku almost succeeds *too well*, as music using her voice often gets described as "by Hatsune Miku" and nobody pays attention to the human who composed the song. If AI gets good enough to compete with top artists but cheap enough to run on a desktop, how do you actually make money off of it?)
There's enough cultural crossover to use their pop media as a successful example, I think. The musical Cats was just as big there as in the West, and manga has thoroughly trounced the American comic industry within America. More specific to pop music, there's the clear trendline from Japanese Idol groups to BTS having 5 songs hit number 1 on the Billboard Hot 100.
I'm not sure I really see Miku and AI music as synonymous, or that the former predicts the latter, but I do think there's precedent for success in Japan turning into success in the Anglosphere.
>I’m kind of confused what’s going on here. Does someone keep throwing in money to push it up to 70-90%, and then other people keep buying it back down?
If this was for real money I'd suspect insider trading. After all, well before an invasion happened PRC generals and probably US/Taiwanese intelligence would know it was going to happen; you have to plan a D-Day-scale event well in advance. And there's a good amphibious window in a couple of months, so it's not incredibly implausible that such a plan is in motion.
...but it's not real money, so insider trading doesn't seem very sensible.
DDG takes its search results from Bing, which has censored searches regarding invasion of Taiwan (just figured that out upon doing my own search).
Anyway, it's mostly to do with weather conditions (typhoons, gales) for transporting the required amounts of materiel (tides change daily, so while they're relevant they don't narrow it down much). I took it from here:
I don't think OSINT would be a long way behind the US and Taiwan in detecting preparations for an invasion. Landing craft are big enough that anybody can check whether thousands of them are being moved to Fujian.
The PLAN doesn't actually have thousands of military landing craft. My (poorly-informed) understanding is that invasion plans involve requisitioning civilian ships, which wouldn't necessarily be obvious (to us).
Has anybody ever figured out how to get thinking going about not widely expected events?
For example, in late August 2015, Angela Merkel opened the borders to a million Muslim refugees, an eye-opening event that helped lead to Brexit and Trump. Were any prediction markets tracking the chances of something like that before it happened, or did it not occur to anybody to ask about it?
Can someone explain how the bets with really long time horizons work?
"Will there be war with China before 2050" won't be resolved for 28 years for example, any money placed on that bet would have an opportunity cost something like 28 years of compound returns at say ~10%, about 14x. You'd need an absolutely huge edge to make that bet worthwhile.
I think the markets work only if they are liquid and there won't be enough liquidity for all markets if you allow arbitrary time horizons and iterated predictions of what other will predict at certain time point.
Or, put the other way, suppose in near-future prediction-market utopia there is significant amounts of money in each of "Will there be war with China before 2050", ".. in 2049", "... in 2048", .... *and* "Will there be war with China before 2023" markets (and by significant amount I mean, enough that market self-corrects to correct probability when people make bad bets). And because we are assuming a prediction market utopia, it would be 28 markets for every question of interest prior to 2050 ... that is going to tie a lot of capital.
In other words, in reality I expect it would not have enough capital, and it would play out like according to principle "the market can stay irrational longer than any given rationalist predictor can stay solvent". Especially if China notices that serious people in the West take "Will there be war with China" markets seriously and realizes it can affect the Western response by burning money in them.
*Ignoring all that*, the opportunity cost problem could be solved if you could trade in instruments like shares in S&P 500 accumulating ETF. Whoever wins is then guaranteed accumulated S&P 500 returns.
It'll resolve early if there is one, though, which biases it upward.
On the other hand, a PRC/USA nuclear war would plausibly result in the prediction market being destroyed, so there's limited ability to collect on a "yes", which biases it downward.
I mean, yeah, obviously. That's what a racketeering organisation *is* - an entity that provides the services of a government (in particular, contract enforcement) to industries that cannot make use of the actual government.
I’m not sure I understand Nostalgebraist’s criticism you quoted. Won’t a “real money” prediction market just have the same biases as the play money ones except now the ability to be an influential player is limited to those who have sufficient money?
I guess the argument I expect here would be that the folks with money would be motivated to discover and fund superforecasters who may not otherwise be able to enter the market at a meaningful level. I’m not sure I totally buy that, since the place to discover a superforecaster would be a play-money market for nerds. If that already exists, then the only purpose of the real money market is to make investors money on a replica of what is already happening in the play-money market, and the quality of the predictions doesn’t improve.
The point of the real-money market is that because it takes real money to move the predictions, people who are bad at prediction get disincentivised from continuing to bet (they lose money) and people who are good at prediction get incentivised to bet a lot (they earn money). You can't make "play a play-money prediction market" a day job that pays for food, but you can with a real-money one.
> Won’t a “real money” prediction market just have the same biases as the play money ones except now the ability to be an influential player is limited to those who have sufficient money?
Only if forecasting is trivial and "superforecasting" is just some kind of genetic luck.
If forecasting is hard – if making a good prediction usually requires research or at least deep thought – then a real money market can compensate for that effort in a way that a play money market can't.
Real money markets also somewhat discourage "trolling" behaviours by imposing a financial cost. In a real money market, "troll" bets are the proverbial $20 on the sidewalk, free to be picked up by any confident forecaster. In a play money market, there's no great incentive to fix a broken market price; there's little advantage to the _market_ side of the prediction market over a mere list of predictions.
I think foxes are more accurate than the hedgehogs in their ecosystem, but a functioning ecosystem needs a bunch of hedgehogs. Hedgehogs are the ones that generate the weird obsessions that foxes can then average out. But a community that only consists of foxes will just get bogged down in groupthink and conventional wisdom because they don’t have anyone pushing their Overton windows.
A lot of people assume foxes are the ones that *break out* of groupthink, but they do this by synthesizing the competing ideas of other groups. It takes a hedgehog to have the obsession with one weird idea to push that makes them see some possibilities, while the fox needs an overview of several hedgehog communities to get the estimates on these possibilities better than any individual community.
PredictIt and Polymarket are out of alignment on the 2024 GOP nomination market because PredictIt is shutting down in February 2023. Everyone now has either pulled their money, or are just gambling on how PredictIt will resolve open markets when they are forced to shut them prematurely.
The problem is that Predictit was shit. Most people on it made money not by holding a bet to resolution, but through volatility. So they were not saying "is this risk priced correctly" but rather "will other people think this risk is priced correctly" or "will future events likely change this price" - which seem similar but are not the same thing. And there was massive amounts of collusion and market manipulation.
Furthermore, the whole idea of prediction markets is generally misguided. The thought is that they serve to aggregate (and perhaps process) information. But if there is no information going in, then there is nothing to process. When there is no clear relationship between informational inputs and desired outputs then all you are measuring is bias and noise.
The best example of this was the Predicit cabinet post markets, which had terrible records at predicting the correct nominee (reasonably far in advance - not two minutes after the nominee's name was leaked to twitter). Predicit would add names as those names were floated in various Politico, WaPo, or NYtimes articles. Grifters would assemble positions and then coordinate rumor campaigns in an attempt to move the price. Some rando nobody expected would be nominated.
The clearest examples of Predictit violating the terms were the Trump tweets markets. How many times the president tweets is not a political event.
On a similar vein, an arbitrary average of the generic ballot polling is also not a political event.
Another pretty clear violation are multiple markets on the same topic (which party will win, who will win, will a specific candidate win).
The obvious solution is to fine Predictit, and then create specific rules about which markets are allowed. All markets outside of that purview are subject to review.
And given how expensive it was to bet - 5% of winnings and 10% of money withdrawn - it was hard to imagine that they were not making a profit. And given how disastrous their handling of election night 2020 was, they were really creating the potential for the regulatory authorities to look bad for not stepping in.
> Also, I wouldn’t have naively expected that, with $34,455 already invested, an extra $100 would almost double the chance, from 14% to 24%. Probably I should poke around at their algorithm some more until this makes sense to me.
This confusion tripped me up as well. Here's the explanation, insofar as I understand it: If you click on the three horizontal dots next to the M$ bet amount, you can open the "market info" panel, where you'll see that while the trade volume on the market is now M$39k (which makes sense given that 856 bets were made), the liquidity pool is only M$666. In other words, lots of people have participated in the market and then sold back their shares (or something), so the actual liquidity is small and any single transaction, buy or sell, can move the market by a lot.
In fact, this exact confusion is why I eventually stopped participating on the site. I'd put M$500 into the ">=50k Chinese Covid deaths by end of 2022" market (https://manifold.markets/MetaculusBot/will-cumulative-reported-deaths-fro), then eventually realized I'd made a bad bet and wanted to cut my losses, only to discover that my shares that I'd bought for M$500 at 70-80% odds were now only worth M$24 if I sold them all, even though the current market probability was still 21%.
Why? Because despite a market volume of M$4.9k, the liquidity pool is only M$105, so if I sell the shares I bought for M$500 I can only get a fraction of that M$105 (or something). Or maybe it would be more accurate to say that the problem is that the liquidity pool for this market is both small and apparently contains ~210 Yes shares vs. only ~26 No shares, so selling Yes shares is worth very little? Hell if I know.
So that's when I stopped using the site. I'd thought prediction markets were a straightforward way to profit from having better information/calibration. But then I learned that if you want to profit, this is ultimately a market, and so understanding the fine print and market mechanics (which I don't care about at all) matters at least as much, and probably much more, than having accurate beliefs.
It's particularly upsetting that the site shows you relatively meaningless information like trade volume, while hiding the important information (like liquidity) in an unnoticeable pop-up.
Thanks for your critique and explanation of what you saw! We do need to do a better job at making liquidity more prominent and understandable, as well as making important markets more liquid.
Limit orders should be a much better way of expressing your beliefs without running into the "slippage" problem you're now seeing - but that's not as obvious to use. Curious, did you notice that was an option at all?
This is an important point that trips up a lot of new users. The key thing you have to look at is not just what the current probability (aka price) is, but what it changes to after your bet - what's happening is that as you buy more YES shares the price of the next YES share on offer increases, so if you buy a lot of YES shares at once then the price can go up a lot. You should only make bets where you think the ending probability/price is still in your favor - otherwise, that means your bet is too big and you need to bet less. This also means that if you accumulated a large number of shares, selling them all at once usually won't work well - but what you can do is sell them gradually over time. (This is just like how trading on financial markets work if you are trading large amounts relative to daily market volume.)
Here's an analogy that I think helps explain why large amounts already bet don't really matter: You can think of the 30k already bet as locked in - the market already matched up each previous yes bettor with a no bettor and locked in their bet at specified odds. When you make a new YES bet, you have to be matched with someone else betting NO - the amount that is already locked in between past bettors (whether it's 30k or 0) makes no difference, it doesn't affect how new bets are made; the only thing that matters is how much people are willing to buy NO right now, which is a function of the liquidity pool or limit orders. People buying and selling shares doesn't increase the amount of liquidity, only adding limit orders or adding to the liquidity pool affects that.
I think it's very true that prediction markets emphasize not only making predictions but also trading, and that doesn't interest everyone. I personally came from mostly non-market-based forecasting and think both are great in different ways. I have been advocating for Manifold to support making bets in a more prediction-centric way - many people like the framework of "bet M$100 on YES", but I also want to be able to say "I think YES is about 70% likely, figure out an appropriate bet amount for me". I think there are many different users who find one or the other more natural, so ideally both options can be supported.
Calling Kalshi the monopoly is kind of rich. PredictIt absolutely had monopolistic pricing and a weird no-action letter no one else could compete with. Kalshi has a first mover advantage but not a monopoly. Anyone can — and will — go through actual regulatory channels to make competitive markets. Markets are a pretty slam dunk case for regulatory rule setting and oversight as well.
Also VPNs are cute, but the issue with non-US market use for US citizens *I*s *R*eally *S*omething else entirely….
Is it illegal to bet on them, or merely illegal to offer them*? If the latter, there's no issue - you can tell the IRS the truth.
*Like how e.g. downloading pirated stuff isn't illegal, but uploading it is. Or how it's illegal to operate a pharmacy without a licence, but if you buy stuff from an illegal pharmacy that's often not illegal (not always, though).
Well even if it isn't illegal, you owe US taxes on your international income. But the site is presumably reporting it to local tax authorities, so you could end up having your profits taxed twice, and/or having to explain to two sets of tax authorities why your income isn't taxable by them. It's painful enough with perfectly legal stock dividends that have tax treaties, doing it with a prediction site feels like a no-win situation to me.
And then of course as prediction markets grow they'll need real liquidity. Try telling the Chief Risk Officer of a reinsurance firm that you can buy stuff from illegal pharmacies legally sometimes (but not always) and see how quickly he says no,
This doesn't work as well as you think - financial markets tend to break when participants are optimizing for the probability of finishing in the top N, especially when N is small. If I were trying to maximize my probability of winning this tournament in a blended pool of rational and irrational agents, I would be:
1) taking way more risk than I would if my goal were simply to maximize my portfolio's returns
2) taking risks that give me different risk exposures from other participants, even if the arithmetic expectation isn't in my favor.
For example, if the true probability of an event happening is 90%, and various market participants bid it up from 50% to 88%, well, I am probably more likely to sell it than buy it at those prices. The intuition is that if the event resolves YES, I will have bought it at the worst price among all the people who bought YES, but if the event resolves NO then I will have sold it at the best price among the people who sold YES.
Requiring people to trade a variety of markets also doesn't do much - it's pretty easy for people to buy YES at 99% on a variety of already-basically-resolved markets to meet this requirement, especially when those markets will formally resolve quickly.
All in all I expect that this competition tests the skill of "how good are you at gaming formal systems" a lot more than it tests the skill of "how good are you at making well calibrated predictions about the world".
Source: I was a quantitative trader for 2 years and they made us play a bunch of games market trading games with solutions like these during training.
In the updated rules post, they say they'll subjectively evaluate betting behavior to decide the winners instead of just looking at profits in order to discourage this.
> I love Manifold as a company but I’m sometimes skeptical of what I see there - there are too many really low-liquidity markets, too many sources of extra play money, too many stupid “the probability on this market will be under 49% on August 31” troll questions, too little desperation keeping people on their toes. The Salem subtournament solves all those problems
I agree that most of those are problems. But it's worth calling out this one:
> too many stupid “the probability on this market will be under 49% on August 31” troll questions
Questions in that genre aren't problems. They're valuable sources of empirical insight into the general area of "Keynesian beauty contests", which are of considerable interest.
"You know a prediction market has really made it when its insane questions about Donald Trump start getting close to zero. Insight has almost made it."
Scott, as one of the internet's premier prediction market evangelists, you should be better than this. Even the market "This market resolves to true at the end of 2022" wouldn't get much lower than 2%. due to the risk-free rate.
PredictIt's New Zealand arm was already shut down years ago when some local anti-money laundering legislation was passed. I've disliked anti-money laundering laws as a concept ever since.
Hey Scott, I've been following and reading your blog for sometime, and I’m fairly interested in prediction markets as well! I am hopeful that regulatory issues on the US are solved and that the projects you mentioned can grow, not just those that work with fake money.
Have you heard about Polkamarkets? I don't recall you ever mentioning them. They are decentralized and autonomous, but the feature I like the most is that any person can create prediction markets on real-life events.
You can also make forecasts, and resolve markets. They’re built on multichain EVM, Moonriver and Moonbeam networks atm.
Not a very big community, they lack liquidity and volume atm but their app has an incredibly ux/ui, it’s very functional tbh. They’re using DeFI and NFTs in an interesting and entertaining way too
Max Martin ( https://en.wikipedia.org/wiki/Max_Martin ) isn't a household name the way the performers he wrote and produced music for are, but the only songwriters that have had more #1 Billboard hits than he did were Paul McCartney and John Lennon, who are kind of a special case...
You should use "it" as a pronoun for organizations rather than "they". Otherwise there's an inconsistency in verb conjugation: "CFTC is" but "they are". I know it's currently in vogue to refer to organizations in plural but it's grammatically incorrect and leads to such jarring inconsistencies.
Actually Predicit’s violations are more serious than I realized. The initially terms are for “contracts consisting of two sub markets regarding binary election outcomes and economic indicators”
> I think Metaculus’ defense would be that their aggregation method lets them weight the guesses of people with good track records more highly
I should have been clearer about this in my post, but I was mostly taking aim at what Metaculus calls the "Community Prediction" -- which is just a (recency-weighted) average of all the predictions, like a literal poll. (As opposed to the "Metaculus Prediction," which is weighted based on track record.)
The reason I focused so much on the Community Prediction is that, when people talk about Metaculus forecasts, they're almost always talking about the Community Prediction. For example, everyone talking about moves in Metaculus graphs is talking about the Community Prediction, since no one can even see the Metaculus Prediction until a question has closed (unless they spend "Tachyons" to get a sneak preview for a particular question).
I agree that doing track-record-weighted aggregation would resolve some of my concerns. But if you ask the question "does Metaculus do track-record-weighted aggregation?", the answer is almost always "not for the numbers we're discussing, no."
Yeah you will have to declare it on the relevant line for illicit income, just like earnings from e.g. drug dealing. It's not tax-exempt.
A couple of weeks ago I ran a betting market predictions game night for the NYC rationalist weekly meetup (all bets should be settleable by the end of the night and the winner is whoever has the most net points). It is (a) a lot of fun and (b) incredibly chaotic, would strongly recommend more people do it. I wrote a post about how I set it up technically and what I'd do to run it smoother here https://shakeddown.wordpress.com/2022/08/04/what-i-learned-about-running-a-betting-market-game-night-contest/
You get some incredibly weird phenomena - I had one question that was "which question will have the most bet on it", which led to two other questions getting a ton of mana sunk into them as the people who'd bet on them tried to make their questions win out. One of these questions was "which answer will get the most points" (the other was "how many light years away is alpha centauri"), which was won by one guy suddenly sinking all his points into the runner-up just before it closed (and since it had a lot of points invested, he ended up winning the whole game by like a 2-1 margin over the runner-up thanks to this. This was also the guy whose day job is a Jane Street trader).
(Also several other fun attempted manipulations. It really is a great creativity generator).
The 55% chance of war with China by 2050 is honestly surprising. I mean, I perceive China as an extremely rational entity ( that is for a government, governments overall are far less rational than people) and I would assume that the disadvantages of waging a war, should be obvious enough that no rational agent would try it.
But then again, this was the assumption with Russia, that Putin would be rational, see that he does not benefit from starting a war, and not do it. And here we are now.
If you model the regime's utility function as not "make china maximally prosperous" but instead just "keep the regime in power", extreme covid policies could be a sort of flypaper strategy to locate and eliminate potential dissidents before they present a real threat to the regime.
I'm sure some Chinese person is going to pop by and tell us this was exactly the strategy employed by some emperor, perhaps even known by a clever animal name. They've got a few thousand years of history to draw on, and there isn't as much new under the sun as we like to think.
The more I read about China, the less it looks like a stereotypical dictatorship. They have transitions of power, successfully, as scheduled. The old guys are happily retired. The selection process for top government positions is arguably more sane than in western democracies, in the sense that it's not 90% a popularity contest and requires actual experience and competence for the job.
Is keeping power a major consideration for the leadership? Well, I'm pretty sure that's true for a supermajority of political systems everywhere - and in democracies it's touted as a feature, not a bug.
The question is if they are "cheating", aka doing things that meta-influence the normal political process. And, at least at a naive look, neither Covid policies nor long term geostrategic goals do that. For a comparative look, mail-in ballots in the last US elections were a significant meta-game move, as they changed the electoral process shortly before elections. AFAIK, China doesn't have anything similar to that - every move they made (good or bad) is only affecting the political process at object level.
What appears rational to Western observers is often countries that do not insist on exporting their particular ideologies to the rest of the world, as the West does.
Xi and Putin have little interest in the internal affairs of other nations, unlike the democracy and human rights-supporting West.
But don’t mistake that lack of principle for rationality. China and Russia put an enormous amount of stock in not being embarassed on the global stage. Both are also dependent on rabidly nationalistic and misinformed publics, and maintain mind-numbing propaganda to that end. Unfortunately for them, the nationalism they invoke also limits their governments’ ability to make unpopular decisions, in some cases more than democracy would. Lastly, Russia and China are kleptocratic states that rely on the support of a variety of elites, either oligarchs or party members. These people are individually rational, but in practice this means they are very corrupt and rely on the state for their wealth.
So no, China is not particularly rational. In fact, I would argue that they very valuelessness that makes people believe Russia and China are somehow more rational than the West is actually a weakness, since they are only able to build alliances of convenience, rather than alliances of shared values. Witness China’s abandonment of Russia compared to Western support for Ukraine--hardly a close Western ally. Pure self-interest is an untrustworthy trait, and tainting that purity with the pride of nationalism hardly makes a state more trustworthy.
> Xi and Putin have little interest in the internal affairs of other nations
Putin appears to be interested in internal affairs of Ukraine, Belarus, Syria, and a few others. Did you mean that he is not interested in internal affairs of nations far away from Russia?
Not really, no. I mean quite exactly he isn’t interested in the internal affairs of other nations. By “internal affairs,” I mean the type of government, the laws, the degree of respect for human rights, the official ideology, and so on. In geopolitical speak, this is a shorthand for a variety of assumptions that come with realism. By “interest,” I mean caring about those facts in themselves. I do not necessarily care about your life merely because I interfere in it for my own self-interest. Psychopaths manipulate and interfere with people, but to say that they care about people is simply wrong. Realism argues that states act psychopathic, always choosing whatever makes them most secure.
Putin’s interest in Belarus and Ukraine is in making them Russian, or Russian statelets. Xi’s interest in Taiwan is in making it a part of the PRC. Neither leader views these nations as separate from their own, regardless of international boundaries. Even if they did view them as separate, their main reason for conquest is not actually some sort of universal principle, but nationalist-fueled revanchism. Russia wants Belarus and Ukraine to be Russia, regardless of the internal politics of those nations.
You mention a potential caveat of:
>Do you mean that he is not interested in the internal affairs of nations far from Russia?
No. Distance is not the relevant factor here. Take Kazakhstan and Pakistan as alternative examples of nations close to Russia and China respectively. Russia cares about how Kazakhstan is governed only insofar as ethnic Russians there are protected and Kazakhstan does not oppose or threaten Russia. Whether Kazakhstan is a democracy or not is entirely irrelevant to them. Whether Kazakhstan follows the teachings of Alexander Dugin is irrelevant to Russia. Pakistan and China are much the same. China could not care less about whether Pakistan is run by a military junta, democratic Islamists, secular elites, or a communist party. What China cares about are the strategic and economic implications for China. These nations are close to China and Russia, but neither country cares about the internal affairs of the other. What matters is solely how what happens in each country affects China or Russia.
Mongolia is perhaps the best example of this. It is a healthy democracy in the midst of the two nastiest regimes of the 21st century, and yet is on good terms with China and Russia. The internal affairs of Mongolia are irrelevant to both nations because its external relations with each are already excellent.
As for Syria, Putin has zero interest in the internal affairs of Syria. It may look that way, but what matters to him is that the regime in charge is friendly with Russia, as the democratic system has assured in Mongolia. The only Russian attachment to the Assad dictatorship itself is instrumental. It is not about how Syria is governed, it is about how Syria relates to Russia. This is an external, geostrategic consideration. So long as whomever governs Syria allows Russia to use it as an airstrip, Russia cares not a whit what happens to the place.
The point is that neither Russia nor China interfere in the internal affairs of other states for ideological reasons, except when they think that those states are actually secessionist members of their own nation. They are not interested in those internal affairs that are truly internal, and lack any geostrategic consequences. Their elites do not worry about the advance of “Xi Jinping Thought” or “making the world safe for Duginism” as the West worries about human rights and democracy.
The best counterexample of this are Israel and Saudi Arabia, excellent Western allies with sub-par and extremely deficient human rights records respectively. Strategically, there is no reason for the West to care about this. But in fact, constant pressure campaigns are applied, the threat of sanctions looms constantly, and international condemnation is common.
Nations like South Africa are forced by Western nations to change their internal affairs (end Apartheid) despite their loyalty to the West if the West finds those affairs too offensive. I think this is obviously good—apartheid was wrong, and it was right to use sanctions to force its end—but it hardly benefitted the West. Indeed, it turned an anti-communist ally into a neutral party (South Africa joined the Non-Aligned Movement in 1994). The Soviet Union once cared about promoting the global struggle for communism, but neither Russia nor China are such ideological projects.
My apologies for the lengthy reply, but it seemed necessary to provide definitions, context, and examples, particularly as you seemed somewhat unfamiliar with common foreign policy jargon.
Thank you, i found this informative.
The problem isn't with China starting war but with West not being interested in preventing such war. By asserting Taiwan independence West makes moves that limit alternatives (like Taiwan being peacefully re-integrated into China - which is still main Chinese plan) and thus potentially making war the only remaining option that maintains China and Taiwan as one country, even with all the costs attached to it.
Cost of war are high but not to the point of "no rational agent would try it" - and it can be choice between "being slowly strangled through other means" vs "short but powerful struggle which will shift battle to more favourable terms even if it is not guaranteed to provide outright victory".
>The problem isn't with China starting war but with West not being interested in preventing such war. By asserting Taiwan independence West makes moves that limit alternatives (like Taiwan being peacefully re-integrated into China - which is still main Chinese plan) and thus potentially making war the only remaining option<
Yeah, that's how I look at the world too: if you don't give in to my demands, I'm forced to fight you... so *you're* the problem!
"Maintains China and Taiwan as one country"? Are you a Chinese propagandist? China and Taiwan are, very clearly, not presently one country. China does not want to maintain the status quo. China would, instead, like to take over Taiwan's wealth and economy and eliminate the freedoms of their citizens and their government institutions.
They are considered to be one country by both countries - that's the status quo. "One China policy" was echo of that reality.
They only disagree on who should have an upper hand in relationship.
Obviously mainland China is stronger economically right now, and thus absent other influences will eventually prevail; West tries to put their hand on the scale and make Taiwanese democracy win due to Western ideological leanings.
This force has to be met with counterforce to preserve previous trajectory - and China certainly seems prepared to employ it.
The Kuomintang (Chiang Kai-shek's party, and the only permitted party in Taiwan until the 90s saw actual democracy) supported one-China (i.e. still claims mainland China as its territory currently occupied by rebels). The Democratic Progressive Party, which is Taiwan's current ruling party, supports Taiwanese independence.
They've so far not had the balls to actually formally declare independence, because the US has told them not to and the PRC has officially and specifically said that it would invade in case of a Taiwanese independence declaration - pissing off the USA while crossing a PRC red line is not pragmatically very wise.
But in the event of an invasion actually occurring anyway, the US would stop discouraging them and the DPP would declare Taiwanese independence.
But until they actually declare independence invasion is highly unlikely and China can continue believing they'll get them eventually in a peaceful way.
So anything that increases chances of Taiwan declaring independence - be that support for Taiwanese independence idea through (previously unheard of) high-level state visits or military support and training similar to Ukraine to increase costs of invasion - increases potential for war.
>But until they actually declare independence invasion is highly unlikely and China can continue believing they'll get them eventually in a peaceful way.
No. The PRC torpedoed that when it revoked Hong Kong's autonomy. The Taiwanese do not want to live under the Party's boot and cannot trust the Party to keep its promises. The only remaining possibilities for reunification are 1) the PRC liberalises and the Party is destroyed, 2) the PRC invades.
The PRC has, indeed, believed for many years that "we'll do it eventually" on a lot of fronts. But if you haven't noticed, Xi seems to have decided that "eventually" is coming around at last.
I'm not going to accuse you of being an enemy agent like REF did; Paul Keating made a similar argument recently and his credentials of "not being a CPC agent" are pretty impeccable. But it's outdated. Yes, they bided their time. Past tense. Now they intend to collect.
It's not quite as simple as "the PRC wakes up one day and decides to start Armageddon".
The scenario looks something like this.
The PRC line has been that "Taiwan is ours and to hell with what the Taiwanese think, we must invade if peaceful reunification isn't happening". The main thing stopping that is US threats of intervention. But the read I get on PRC thinking is that they really do not get the Western moralistic mindset, so I'd expect their guess at the chance of intervention to be significantly lower than what it actually is (and there's always the fascist overconfidence to consider). So there should be a lot of probability mass on "they attempt an invasion and the West does intervene".
Yes. Honestly, this is just a very real application of what Scott has called "the principle of charity," and if you look for it, you can see it everywhere: not taking moralistic claims seriously. The more cynical you are already, the more prone you are to project that everyone everywhere operates on cynicism. But even not-so-cynical people are vulnerable to this bias if they're sufficiently partisan when thinking about their enemies.
The two most likely triggers for an invasion I can see are this: Taiwan does something that makes the PRC feel forced to act and roll the dice (i.e., formal declaration of independence). Or the US is seemingly paralyzed by a constitutional crisis (most likely centered around an election) and PRC feels a narrow window is opened for action.
The latter could work, but it also seems ripe for a miscalculation in which the US ability to act is underestimated. Kerensky's Russian Republic was somehow able to more or less continue fighting WW1 (albeit ineptly) for several months despite the fact that Russia for all intents and purposes did not have a functioning government at the time. Something tells me the US military + intelligence machinery could still act against China for a period of time despite the US Federal Government being apparently paralyzed, and by the time the US military was running out of steam and needed to requisition more funds, the crisis might well be resolved and the government functioning again.
The US threat of intervention might no longer be the deciding factor in hypothetical invasion scenarios. Air supremacy and naval supremacy had a good historical run, but current warfare (including the one in Ukraine) is heavily reliant on missile supremacy.
Taiwan is (even on its own) lavishly equipped with missiles that can stop China's naval advances, just as China is lavishly equipped with missiles that can stop US Navy advances. Any attempt at a direct invasion of Taiwan would be an unmitigated bloodbath, with or without the US.
Some data points to consider are laid out in https://edition.cnn.com/2022/05/31/asia/china-taiwan-invasion-scenarios-analysis-intl-hnk-ml/index.html
The threat of Western intervention does close off a number of otherwise-effective options the PLA has for escalation.
I'm curious about what you're proposing. A blockade? Or perhaps just laying waste to the place?
The thought I've had is that if the PRC is largely interested in Taiwan for economic reasons, it doesn't require much Taiwanese resistance at all to turn the invasion into a pyrrhic victory. It requires Taiwan to surrender rapidly and China to be very surgical in its bombardment.
But if China's interests are more a matter of prestige and making an example of the place -- costs be damned -- then it can safely lay waste to the civilian infrastructure and starve the island into submission, unless the US is prepared to break the blockade (which was touched on in that CNN article). In this scenario, China might not even be interested in conquering Taiwan -- simply bullying it into vassalage would be sufficient. Let Taiwan pick up most of the bill for rebuilding its own infrastructure.
When it comes to China attempting a sudden, swift conquest, the most important contribution of the US might be to morale. There has been a lot of doubt cast on the reliability of Taiwan's military and reserves, though I think everyone has revised their appraisals upwards after witnessing Ukrainian resistance. Still, the perception of whether or not the cavalry is coming could prove decisive to the marginally motivated ROC soldier.
The PRC is not largely interested in Taiwan for economic reasons. They were willing to wage bloody war over Taiwan when Taiwan was an economic basket case, and the reasons for that have not changed.
Both.
The CNN article misses something re: blockade. Sure, the USA wouldn't go in shooting just because of a blockade - but they wouldn't have to. They'd just *run* the blockade, and throw the dare back in the PLA's teeth with "are you going to shoot at USN ships in waters not recognised as yours". Similar to the "freedom of navigation" missions they've been doing for years - they sail through what the US says are international waters and the PRC claims as territorial waters, and basically dare the PLA/PLAN to do anything about it. Or, for that matter, similar to the Berlin Airlift.
And yeah, obviously the PLARF could nuke the island into oblivion if taking the industry intact was no concern and the West responding in kind was no concern.
"Running" the blockade /does/ mean shooting at ships, not making rude gestures at them as they pass by. The blockade *is* military warfare, not economic one.
As that CNN's comment says, "What's cheap is a surface-to-ship missile, what's expensive is a ship." PRC can easily do "surface-to-ship" to any Taiwanese ship (or plane), and so can Taiwan to any PRC ships in its vicinity. Taiwan can't do much surface-to-ship stuff to PRC ships departing from ports adjacent to Vietnam or Korea. And it's clear that surface-to-city stuff done by either side would be immediately reciprocated by the other. Every military war has escalation levels, and that escalation wouldn't be taken lightly.
The West shooting at PRC ships /would/ mean engaging in military warfare, of course. I don't see that happening; it would remain on the level of economic and media warfare.
A military-enforced blockade of Taiwan is a much more sane option for PRC than a military invasion. By "threat of Western intervention" do you mean that the West would actually declare a (military) war on PRC if PRC blockaded Taiwan?
Nah. Simply no way they would, not over a blockade. But they certainly would escalate the *economic* war. (whereas the West's *media* war on PRC and Russia had already been amped up to 11 for at least the last dozen years, there's not much left to escalate on that front anyway)
I think the threat of this economic war *is* what's effectively keeping PRC at bay, for now. The time is not right. The US dollar has had a virtually monopolistic stranglehold on world trade in the post-WWII era. But it seems this is slowly beginning to change. The US sanctions on Russia have caused the first international oil trades to be settled in yuans instead of in dollars - a watershed moment in that area. Biden's money printing is also not helping the dollar's popularity abroad. When those cracks have expanded sufficiently, then PRC might decide that the time is right to endure an economic war with the West.
See my response to Wency.
China has had a fairly rational approach to foreign policy in the post Deng era but now with oligarchy being replaced by dictatorship we should expect less of that as the whims of particular leaders can have more play.
Deng wanted Superpower China too. He just said "bide your time until the PRC is strong enough, because it isn't yet". Now, 30 years later, well, it's quite a bit stronger and so they're thinking about getting on with it.
Unlike Putin's invasion of Ukraine, which was unambiguously illegal under international law, Taiwan is legally part of the PRC and therefore they have the right to invade it whenever they want, just like the USA has the right to stick as many troops in Hawaii as it feels like.
Also, you are looking at the calculus for China invading Taiwan today. Right now, the USA unambiguously has a stronger navy and its sanctions could easily be crippling, but China is developing quite rapidly economically. Say it's 2040, China and the USA both have domestic semiconductor industries as good as Taiwan's, China has an economy the size of the EU and US put together, and China's navy can compete with the USA in the Pacific. Is war still a drastic miscalculation?
I'd argue that mainland China is a part of Taiwan and Taiwan is justified in reclaiming the mainland any time they feel they're able to.
Sure, that would also be fine under international law, in the same way we've adjudicated countless civil wars. However, unlike the PRC invading Taiwan, it has about a 0% chance of happening.
Has Jeb!'s golden sword been whispering to you too?
Did Jeb! want to liberate the PRC?
This is apparently a reference to the Sword of Chang, which is conjectured to be a confused reference to the idea of Chiang Kai-Shek reclaiming mainland China. Scott mentioned it (with some links to articles) here:
https://slatestarcodex.com/2015/11/16/hardball-questions-for-the-next-debate/
>Taiwan is legally part of the PRC and therefore they have the right to invade it whenever they want
At these scales there's no such thing as de jure, only de facto.
While the law is clearly much less effective in international relations than power, I think it is also too cynical to suggest it means literally nothing. There is a reason why the PRC obsessively polices people's adherence to the One China principle.
China's government uses CCP ideology as a co-ordinating glue that holds everything together and reduces the odds of late Soviet terminal ennui. That co-ordinating glue has as one of its central planks "reclaim Taiwan by 2049".
This clock was set ticking by Mao talking to Nixon:
https://chinafactor.news/2021/11/22/taiwan-can-wait-100-years-mao-told-nixon/
And has been re-iterated many times since.
Mostly recently:
https://www.newsweek.com/china-unify-taiwan-within-30-years-analyst-1585279
More dramatically, there's a story (whose link I can't find right now) about a high-ranking US official from the Clinton (or Obama?) administration having a chat with with a high-ranking CCP official at the sidelines of some conference. The US official casually said "And Taiwan is still a thing we can punt for 100 years right?" and the CCP official replied "No, that was X years ago, so now it's only 100-X years"
At the time, America interpreted 100 years as "forever away" and the CCP interpreted 100 as "actually 100". Given that the CCP used the terms of a 99 year treaty to reclaim Hong Kong in 1997, this is relevant.
I think calling "more than 100 causalities" a war is stretching it. I would call it a skirmish. You could easily get to that number by some naval ship crossing into territorial waters and then being sunk.
Even a conventional war would probably quickly rise in the tens of thousands of death, at least.
Vicariously Predicting Noncompliance
Voluntarily Performing Nobly
Have not read it all yet but have stopped to View Products Not-named. This will help me. Thanks.
The 10th Avatar of Vishnu is Kalki, not Kalshi.
Exactly! Kalshi supposedly means 'everything' in Arabic. Please correct asap.
Given that when one hovers over the picture, the pop-up text includes a title that identifies the portrayed figure as "Kalki", it seems likely that Scott is aware of this.
Which is why I chuckled at what seemed to me to be a joke based on Scott purposefully making fun of the American tendency to confuse foreign words, while engaging in a little bit of hyperbole associating Kalshi with apocalyptic things.
Ah - I had not seen the popup. Thanks!
M$1 on Manifold is one cent USD, so when you say there was $4000 bet on the AI company data breach question, that's actually USD $40.
Re "I’m kind of confused what’s going on here", I would guess that while there's a lot invested, there's not many limit orders on the 'no invasion' side (because China might invade Taiwan any day, or there might be news that it's suddenly much more likely to do so, and then the probability spikes for real so your limit order is filled but it's now a bad investment). You don't want to leave open an offer on 'no invasion', but you'd happily buy 'no' if you see that the price is way too high. Thus one buy on 'yes' can change the probability a lot.
Does anyone else enjoy predicting what would happen in fictional worlds is the laws of narrative and plot were suddenly suspended? Like Palpatine says "execute order 66" and then suddenly Star Wars stops being a narrative universe and just proceeds in a way that is dictated by its own internal laws, but no tendency towards narrative arc- what happens?
The difficulty is that in some fantasy universes, the narrativity is written into the plot. In Star Wars, very explicitly and in Dune, implicitly, what happens is the outworking of a superhuman intellectual agency.
Still, it's a fun exercise.
The most fun to be had as a dungeon master, I think, is to make sure the player characters have concrete motivations, give the NPCs sometimes-aligning, sometimes-opposing motivations, set a bunch of internal laws about what strategies the NPCs will follow, and just simulate the world as accurately as possible, without regard for plot or narrative arc. Doing it that way actually produces a far more interesting, satisfying and unexpected narrative arc than what I would have written, in my experience.
I think this is because the human mind has an amazing capacity to find meaningful patterns in events, and even when the events are randomly-generated or accidental, those patterns will appear on their own and start exerting a gravitational pull on everything.
I once (at age 21?) ruined (or made *amazing*) a 12? hour DnD session. We were captured less than an hour into it, and I spent the entire time trying to escape a jail cell when the DM stated some logical inconsistencies and refused to back down from them. Many laughs were had.
Specifically we were served food through a slot in our cell door, but no one was emptying out toilet buckets. But the bucket were emptied...
So there were hours pushing the DM about how the magic shit buckets worked.
In 2016 on PredictIt, I saw one case, which seems like undeniable insider trading. I was on the RealClearPolitics poll market, and someone was buying large amounts of a 2c YES stock, 10 minutes minutes before closing (around midnight EST). A few minutes later, a new poll was added to RealClearPolitics, causing the correct answer to flip. Even with the $850 market cap, buying in at 2-5c, would mean $20k+ in winnings.
I can't speak to how commonly this occurred.
In prediction markets is insider trading a bad thing? I thought the whole point was to produce accurate predictions by incentivizing people to use any accurate knowledge they had. Sounds like your insider trader was moving the market in the correct direction. Or was the 538 post inaccurate?
It's a hard and complicated question. My intuition (with no policy implications) is we want markets to incentivize doing research in advance. An insider finding out the answer two seconds before the market closes when info gets publicly released provides almost no value and invalidates the expected return of all those who put in research adding liquidity to the market in the first place.
> [A]s far as I can tell this is only about who makes it to the finalist round, and then to some degree the usual criteria will start creeping in again.
Yeah, that's what it says.
I've been doing some of this prediction market stuff and have been pretty good at it. (I only predict things I have domain knowledge about. But I'm told that's not an issue.) And this sort of thing keeps me out. It keeps me out of a fair number of things where I know there's going to be the same old standard filter applied after all the fancy new ways of sorting. Layering an additional test on top of the standard one only selects for people already in the circle. So if I'm not in I expect I can beat them all and still lose on points to the Harvard PhD.
Another way to think of it: Imagine a normal fellowship takes 100 applications and selects 1 scholar based on normal academic criteria. They disqualify 60 of them and I'm in the 60 because I don't have a PhD. But they chose one of the 40. I won't apply. Imagine this fellowship takes 100 applications and uses prediction markets to narrow it down to 5 scholars. And then selects based on normal academic criteria. Even if I'm 100% confident I can get into the top five, even I'm 100% confident I will be number 1, my expectation is that I get into the top five and then don't get the scholarship. So my expectation is that both of these contests end with basically the same result. The only thing that shifts is possibly which of the 40 who would make it through the original filter would win. The only way it won't is if all 5 winners are not people otherwise qualified which is unlikely. That's something but not enough for me to feel like I'm actually in the running.
The net filter is at least somewhat equivalent to asking people for resumes, selecting the qualified ones and excluding everyone else, and then having them do a prediction market to decide who wins.
This is without getting into the tournament criticisms here which I broadly agree with:
https://www.cspicenter.com/p/introducing-the-salemcspi-forecasting/comment/8235817
PS: To be explicit I have no idea if I'd win. But my strong sense is that even if I did win I wouldn't get beyond being a finalist.
Do you think this provides a way to pad your conventional resume with your unconventional prediction market skills? ("Your" is meant to be generic, I have no idea if you personally are interested in that sort of thing.) Superficially "I was officially recognized as one of the finalists for this cool fellowship" sounds like a resume-worthy accomplishment -- but maybe anyone who'd care would already know that the fellowship in question is kind of weird and therefore not care.
I do not. Firstly, I expect most people to not know what this is and in the worst case to see it as me bragging about being a good gambler. And secondly I expect most other rat-adjacent jobs to have the exact same issue as this contest. They're not searching for the best predictor. They're searching for the best predictor who's a graduate student of an elite college or similarly impressive.
In this the rationalist movement seems significantly worse at finding underappreciated talent than several large corporations let alone some more insurgent movements. Instead its primary ability is to attract a certain kind of high class talent who'd otherwise work at Deloitte. Which isn't nothing but certainly isn't so much rationally getting past credentialism as running a hype train.
Isn't "bragging" what resumes are for? (You know much more than I about rat-adjacent hiring practices -- I'm both credentialed and not rat-adjacent -- so it's interesting to learn about.)
Sorry, emphasize "about gambling" not "bragging."
The filter is different if none of the top 5 have the qualifications though. And 5 isn't too large to imagine that happening.
I mentioned this. It's not impossible but it's unlikely. And even if I get through the question is then whether I'm the best of the five and so relies on my estimation of the median credentials of who's involved. Which I expect to be generally better than mine.
Same thought here: why bother? This is hardly any different than a regular application process, at least insofar as "only apply if you think your resume is good enough to get the job" — how you do on the prediction competition isn't really a factor unless you *already* believe you're a likely hire.
Thanks for registering your agreement. Makes me feel less like a lonely curmudgeon.
Slightly OT, but a more successful job search strategy (outside academia) for me has been to apply if I think the chance of getting *an interview* is worth the effort required (e.g., "easy apply" roles are less effort, so fire away). Your chance of getting the job, conditional on getting an interview, is largely independent of your resume.
Yeah I was disappointed to see the format of the tournament, likely no more predicative than a March Madness Pool, where that one random fan who loves terrible team X wins for whichever team X surprises.
If the MATH dataset is open source, is there no guarantee people aren't just overfitting the test set to meaninglessly pad their accuracy in ways that won't generalize out of sample? To run a real ML competition you would need a secret test set that the participants never get access to.
It looks like the full test set is available here: https://www.kaggle.com/mathurinache/math-dataset
So I'm tentatively suspecting that overfitting is the more likely explanation than AI capabilities proceeding 7SD faster than the forecasters expected. It is trivial to just memorize the answers to the test set.
ML benchmarks are often run with public test sets, so yeah it's sorta in the honor system. On the other hand, it's been demonstrated that while the community has overfit e.g. CIFAR10's test set in that the accuracy will drop if you resample the data, higher performance on the CIFAR10 test set still predicts higher performance on imagenet (of the same architecture, not the exact same weights).
If you think about it, you get such slow feedback from the test set and it's so few bits that it's really hard to imagine actually having your model implicitly memorize examples through hyperparameter settings.
> I still think the hedgehog was supposed to be the bad one.
Whatever you say Robotnik
People reading this might come away with the impression that PredictIt was still mostly in compliance with the no action letter, since Scott chose to only highlight three of the nine requirements in the no action letter. Just to be clear, and as I suspect Scott is aware, PredictIt was in substantial violation of a number of the other requirements in the no action letter, such as the requirement to keep the subject matter of the markets to a narrow range of subjects and limits on overall revenue generation.
Which markets were outside the subject range? I see only politics and a couple of economics markets on PredictIt. The original letter listed said it would "include" three specific things in politics, but given that PredictIt has been doing more than that forever I assumed there was an understanding that they could branch beyond that as long as they stayed within politics.
What was the limit on overall revenue generation? I mentioned the issue with Aristotle being for-profit, I don't see any other limits (eg specific dollar amounts) mentioned in the letter.
I am not really convinced they're violating the letter, especially because if they wanted to violate the letter there would be much more profitable ways for them to do so (eg removing money limits, removing trader limits) plus they made some changes (like closing their popular tweet markets) that I interpret as regulators telling them to do things and them listening. Probably this involved some limit-testing and limit-pushing, with them taking CFTC not caring as encouragement. My model is that they have tried to comply, have been modulating their exact compliance level based on interactions with CFTC over the past decade, thought they were in compliance with the law-as-it-really-exists-and-is-enforced, and this came as a surprise to them.
PredictIt was certainly aggressive with the $850 limit - they'd routinely list 2+ virtually identical contracts each with independent $850 limits. But as you say they've been doing that forever.
I am surprised about the question regarding pop songs being composed by AI; that is, I was assuming that they were already mostly composed by AI. Was I wrong ? Are there really a bunch of human composers still toiling in the pop mines -- or are humans just sitting at the top of the loop, clicking "next step" when they find a generated song that they like ?
Most pop songs are definitely made by people at the moment lol.
It is complicated, the person’s whose name is on the writing credit may have bought samples from several people and combined them into one tune, but music is made by actual humans.
AI music is currently rather poor, especially if it includes human vocals or needs to have human vocals put over it.
My reaction was "they expect there to still be a Billboard Top 100 by 2050?" Music is already splintered into a thousand sub-subgenres; by 2050 Son-of-Spotify will be producing personalised DJ mixes for everyone based on their mood and biosigns, or even totally novel personalised music. But I guess institutions linger, and there will probably still *be* a Billboard Top 100 even if hardly anyone listens to the songs on it.
I think some people like the fact that they listen to the same music other people do. The specific mechanism in the future may be quite different from now... maybe all music will start as a personalized production, but then people will share it on social networks, and you can have a list of "songs that received most likes"... but it still may be some form of Top list.
Billboard has been integrating streaming music stats in its charts for a long time now. And while there are subgenres, there still doubtlessly is the mainstream, particularly relevant to people who aren't music enthusuiasts, with widely known smash-hits still emerging regularly.
True, but a kid born today will be 28 in 2050. The generation that will be defining musical tastes in the 2050s will have known streaming and algorithmic recommendations all their lives, and think of them as normal. They'll probably have a very different relationship to music than today's tastemakers.
Plausible, but I would expect there to still be tastemakers of some sort. Algorithmic recommendations aren't arbitrary, to be effective they have to reflect underlying human nature, with its herd behavior tendencies.
True, but kids born in the late 90s and early 2000s are listening to pop music that's shockingly similar to what was being made when they were born.
Here's a fun, depressing lecture about it:
https://www.youtube.com/watch?v=aCgkLICTskQ
I always thought I was alone in thinking "basically all of the pop music after around the year 2000 sounds like it could have come out in the year 2000". Turns out it's at least a fringe or minority opinion.
Maybe pop music simply achieved its long sought holy grail of lowest common denominator around that time, and there's no more "improvement" to be had? An insightful video along these lines:
https://www.youtube.com/watch?v=K0Vn9V-tRCo
Very nice. Gotta admit, I was expecting (and slightly looking forward to) a rickroll.
Update: I got around to watching this and it was great.
Here, have an 8 bit Master of Puppets in return: https://www.youtube.com/watch?v=66Bmumvw3NM
I think a challenge for AI-created music is that there is already an oversupply of music, because lifetimes' worth of music have already been recorded and can be listened to instantly, and a large number of extremely talented people are working for pennies to record music faster than it can be listened to. Next time you're in Nashville, ask your waiter if he's on Spotify.
My first thought is that AI-generated music, to be useful, would need to be applied to very-specific applications that no actual musician will touch, at least not for a price that works. For example, suppose I'm trying to study for an exam, and there was software that could transform information I need to memorize into a catchy jingle? Or my factory wants to do the same thing regarding its safety procedures?
Are there any short term AI forecasting benchmarks I can make predictions on to see how well calibrated I am?
Why didn't predictit:
1. spend some of their zillions of fees on hiring former CFTC people and lobbying the CFTC to let them continue existing?
2. Make a big public fuss about the CFTC's corruption?
3. Move their operations outside the US and stop accepting US customers due to their Vexing Political Nannystate?
I am confused why they didn't do any of this, unless they are planning to relaunch the site under a different legal framework.
1. I don't think they're actually making that much profit on fees.
2. They are trying to do this.
3. As currently designed most of their contracts are on US politics; I think "foreigners who want to bet on the US government and aren't already using other foreign prediction markets" is a pretty small market.
But they still have six months until they're shut down, so maybe they will do one of these things.
I don't think that trading has stoped on PredictIt.
You're right, thanks. When I checked over the weekend all trading it was suspended, but it looks like that was just for routine maintenance.
Netpick, but UTA is the University of Texas Arlington. UT Austin is just UT.
The Austin campus is renowned for the courage and attractiveness of its students.
The Arlington campus is know for its students not being able to get unto UT.
The Dallas campus is known for their students not being able to get into UTA.
UTDallas' average SAT scores are consistently ~150 points higher than UTArlington's, and are competitive with and occasionally higher than UTAustin's. Source: https://data.utsystem.edu/data-index/admissions-test-scores
I dated a Comet back in the early 2000s. A brief fling with a Maverick a few years later. My references may be dated.
So, Ravenclaw, Gryffindor, Hufflepuff?
"This seems underpriced, unless people expect whatever group manages the list to disqualify AI songs, or some kind of civilization-wide agreement that they don’t count. There’s already mediocre AI-produced music, and nothing stopping it from getting better with really simple scaling. Given that most songs are already composed by nameless corporate composers and not the pop stars themselves, I don’t think there will be any resistance to this."
Uh....
Pop music isn't what's best, it's what's pushed. It has to be good, but more importantly it has to be pushed, and it has to have the kind of cachet to be pushed.
So this is essentially a bet that the cool crowd of cool kids will decide it's cool to push computers as artists instead of exclusively pushing other (relatively talented and convincing and possibly interestingly insane) cool kids. Will "this was made by a computer, oooooh" be competitive with staring into the absurdly compelling eyes of the likes of Taylor Swift and Billie Eilish and Adam Levine?
You know I'm down on AI, but 43% seems perfectly priced to me.
This isn't about performers, though, it's about composers. The current #1, _Break My Soul_ is promoted as "by Beyoncé", but she was one of four songwriters who collaborated on the song, two of whom I've never heard of. Same deal with #2 - you have to go down to #4 before you find a song written solely by the featured artist (Kate Bush), and the next one is at #9 (Nicky Youre): https://www.billboard.com/charts/hot-100/ So I think it's likely that we'll see AI *composed* music long before we see Idoru-style *performers* - the audience will neither know nor care that North West's latest banger was written by an AI.
Random, but since you pulled me back to 80s music and your name is Chaz, my brain really tried to convince me that you were Chaz Jankel, writer of "Number One" and other 80s fare.
I laughed at Bugmaster's comment too.
Yes, this one: https://www.youtube.com/watch?v=-aLYvZ5sX28
(I actually like Taylor Swift alright)
That's true! If the condition of the bets allow partial or full credit for writers as distinct from top of line performers, then yeah, this could be undervalued.
Hatsune Miku is evidence that you can push a computer as a pop artist, although getting 50% of the Billboard top charts is a very high bar to clear.
(Actually, Miku almost succeeds *too well*, as music using her voice often gets described as "by Hatsune Miku" and nobody pays attention to the human who composed the song. If AI gets good enough to compete with top artists but cheap enough to run on a desktop, how do you actually make money off of it?)
Hatsune Miku is Japan and that's a whole different prediction.
There's enough cultural crossover to use their pop media as a successful example, I think. The musical Cats was just as big there as in the West, and manga has thoroughly trounced the American comic industry within America. More specific to pop music, there's the clear trendline from Japanese Idol groups to BTS having 5 songs hit number 1 on the Billboard Hot 100.
I'm not sure I really see Miku and AI music as synonymous, or that the former predicts the latter, but I do think there's precedent for success in Japan turning into success in the Anglosphere.
Minor quibble:
> Also, isn’t Kalshi the apocalypse-avatar of Vishnu who’s supposed to kill everybody and purify the world for the next era? Seems concerning.
That's _Kalki_
Pretty sure that the purposeful confusion of the names is a purposeful joke.
At least, I chuckled. And I'd be willing to bet that Scott knows Hindu religious figures better than I do.
Either I didn't get the joke, or he made a mistake coz I am pretty sure I know my Hindu gods better than he does
I think the Predictit 2024 gop primary market will probably never resolve which would ruin the arbitrage.
>I’m kind of confused what’s going on here. Does someone keep throwing in money to push it up to 70-90%, and then other people keep buying it back down?
If this was for real money I'd suspect insider trading. After all, well before an invasion happened PRC generals and probably US/Taiwanese intelligence would know it was going to happen; you have to plan a D-Day-scale event well in advance. And there's a good amphibious window in a couple of months, so it's not incredibly implausible that such a plan is in motion.
...but it's not real money, so insider trading doesn't seem very sensible.
What determines an amphibious window? Tides? A quick DDGo search isn't turning up anything. I'm 50/50 that it's a joke I didn't get.
DDG takes its search results from Bing, which has censored searches regarding invasion of Taiwan (just figured that out upon doing my own search).
Anyway, it's mostly to do with weather conditions (typhoons, gales) for transporting the required amounts of materiel (tides change daily, so while they're relevant they don't narrow it down much). I took it from here:
https://foreignpolicy.com/2018/09/25/taiwan-can-win-a-war-with-china/
Is it possible they might invade at some other time? Sure. But April and October seem the best bets.
Fascinating, thanks.
I don't think OSINT would be a long way behind the US and Taiwan in detecting preparations for an invasion. Landing craft are big enough that anybody can check whether thousands of them are being moved to Fujian.
The PLAN doesn't actually have thousands of military landing craft. My (poorly-informed) understanding is that invasion plans involve requisitioning civilian ships, which wouldn't necessarily be obvious (to us).
Has anybody ever figured out how to get thinking going about not widely expected events?
For example, in late August 2015, Angela Merkel opened the borders to a million Muslim refugees, an eye-opening event that helped lead to Brexit and Trump. Were any prediction markets tracking the chances of something like that before it happened, or did it not occur to anybody to ask about it?
Can someone explain how the bets with really long time horizons work?
"Will there be war with China before 2050" won't be resolved for 28 years for example, any money placed on that bet would have an opportunity cost something like 28 years of compound returns at say ~10%, about 14x. You'd need an absolutely huge edge to make that bet worthwhile.
He talks about it in this one: https://astralcodexten.substack.com/p/mantic-monday-1115 . Both "reciprocal markets" and "iterated markets" sections discuss this issue.
(And the "reciprocal markets" sections references the beginning of this prior post: https://astralcodexten.substack.com/p/mantic-monday-11121 )
*reciprocal scoring *iterated prediction markets
I think the markets work only if they are liquid and there won't be enough liquidity for all markets if you allow arbitrary time horizons and iterated predictions of what other will predict at certain time point.
Or, put the other way, suppose in near-future prediction-market utopia there is significant amounts of money in each of "Will there be war with China before 2050", ".. in 2049", "... in 2048", .... *and* "Will there be war with China before 2023" markets (and by significant amount I mean, enough that market self-corrects to correct probability when people make bad bets). And because we are assuming a prediction market utopia, it would be 28 markets for every question of interest prior to 2050 ... that is going to tie a lot of capital.
In other words, in reality I expect it would not have enough capital, and it would play out like according to principle "the market can stay irrational longer than any given rationalist predictor can stay solvent". Especially if China notices that serious people in the West take "Will there be war with China" markets seriously and realizes it can affect the Western response by burning money in them.
*Ignoring all that*, the opportunity cost problem could be solved if you could trade in instruments like shares in S&P 500 accumulating ETF. Whoever wins is then guaranteed accumulated S&P 500 returns.
It'll resolve early if there is one, though, which biases it upward.
On the other hand, a PRC/USA nuclear war would plausibly result in the prediction market being destroyed, so there's limited ability to collect on a "yes", which biases it downward.
More evidence that governments are organized crime rings with better PR.
I mean, yeah, obviously. That's what a racketeering organisation *is* - an entity that provides the services of a government (in particular, contract enforcement) to industries that cannot make use of the actual government.
I’m not sure I understand Nostalgebraist’s criticism you quoted. Won’t a “real money” prediction market just have the same biases as the play money ones except now the ability to be an influential player is limited to those who have sufficient money?
I guess the argument I expect here would be that the folks with money would be motivated to discover and fund superforecasters who may not otherwise be able to enter the market at a meaningful level. I’m not sure I totally buy that, since the place to discover a superforecaster would be a play-money market for nerds. If that already exists, then the only purpose of the real money market is to make investors money on a replica of what is already happening in the play-money market, and the quality of the predictions doesn’t improve.
The point of the real-money market is that because it takes real money to move the predictions, people who are bad at prediction get disincentivised from continuing to bet (they lose money) and people who are good at prediction get incentivised to bet a lot (they earn money). You can't make "play a play-money prediction market" a day job that pays for food, but you can with a real-money one.
> Won’t a “real money” prediction market just have the same biases as the play money ones except now the ability to be an influential player is limited to those who have sufficient money?
Only if forecasting is trivial and "superforecasting" is just some kind of genetic luck.
If forecasting is hard – if making a good prediction usually requires research or at least deep thought – then a real money market can compensate for that effort in a way that a play money market can't.
Real money markets also somewhat discourage "trolling" behaviours by imposing a financial cost. In a real money market, "troll" bets are the proverbial $20 on the sidewalk, free to be picked up by any confident forecaster. In a play money market, there's no great incentive to fix a broken market price; there's little advantage to the _market_ side of the prediction market over a mere list of predictions.
On hedgehogs and “good” or “bad”:
I think foxes are more accurate than the hedgehogs in their ecosystem, but a functioning ecosystem needs a bunch of hedgehogs. Hedgehogs are the ones that generate the weird obsessions that foxes can then average out. But a community that only consists of foxes will just get bogged down in groupthink and conventional wisdom because they don’t have anyone pushing their Overton windows.
A lot of people assume foxes are the ones that *break out* of groupthink, but they do this by synthesizing the competing ideas of other groups. It takes a hedgehog to have the obsession with one weird idea to push that makes them see some possibilities, while the fox needs an overview of several hedgehog communities to get the estimates on these possibilities better than any individual community.
This is well-put.
PredictIt and Polymarket are out of alignment on the 2024 GOP nomination market because PredictIt is shutting down in February 2023. Everyone now has either pulled their money, or are just gambling on how PredictIt will resolve open markets when they are forced to shut them prematurely.
The problem is that Predictit was shit. Most people on it made money not by holding a bet to resolution, but through volatility. So they were not saying "is this risk priced correctly" but rather "will other people think this risk is priced correctly" or "will future events likely change this price" - which seem similar but are not the same thing. And there was massive amounts of collusion and market manipulation.
Furthermore, the whole idea of prediction markets is generally misguided. The thought is that they serve to aggregate (and perhaps process) information. But if there is no information going in, then there is nothing to process. When there is no clear relationship between informational inputs and desired outputs then all you are measuring is bias and noise.
The best example of this was the Predicit cabinet post markets, which had terrible records at predicting the correct nominee (reasonably far in advance - not two minutes after the nominee's name was leaked to twitter). Predicit would add names as those names were floated in various Politico, WaPo, or NYtimes articles. Grifters would assemble positions and then coordinate rumor campaigns in an attempt to move the price. Some rando nobody expected would be nominated.
The clearest examples of Predictit violating the terms were the Trump tweets markets. How many times the president tweets is not a political event.
On a similar vein, an arbitrary average of the generic ballot polling is also not a political event.
Another pretty clear violation are multiple markets on the same topic (which party will win, who will win, will a specific candidate win).
The obvious solution is to fine Predictit, and then create specific rules about which markets are allowed. All markets outside of that purview are subject to review.
And given how expensive it was to bet - 5% of winnings and 10% of money withdrawn - it was hard to imagine that they were not making a profit. And given how disastrous their handling of election night 2020 was, they were really creating the potential for the regulatory authorities to look bad for not stepping in.
> Also, I wouldn’t have naively expected that, with $34,455 already invested, an extra $100 would almost double the chance, from 14% to 24%. Probably I should poke around at their algorithm some more until this makes sense to me.
This confusion tripped me up as well. Here's the explanation, insofar as I understand it: If you click on the three horizontal dots next to the M$ bet amount, you can open the "market info" panel, where you'll see that while the trade volume on the market is now M$39k (which makes sense given that 856 bets were made), the liquidity pool is only M$666. In other words, lots of people have participated in the market and then sold back their shares (or something), so the actual liquidity is small and any single transaction, buy or sell, can move the market by a lot.
In fact, this exact confusion is why I eventually stopped participating on the site. I'd put M$500 into the ">=50k Chinese Covid deaths by end of 2022" market (https://manifold.markets/MetaculusBot/will-cumulative-reported-deaths-fro), then eventually realized I'd made a bad bet and wanted to cut my losses, only to discover that my shares that I'd bought for M$500 at 70-80% odds were now only worth M$24 if I sold them all, even though the current market probability was still 21%.
Why? Because despite a market volume of M$4.9k, the liquidity pool is only M$105, so if I sell the shares I bought for M$500 I can only get a fraction of that M$105 (or something). Or maybe it would be more accurate to say that the problem is that the liquidity pool for this market is both small and apparently contains ~210 Yes shares vs. only ~26 No shares, so selling Yes shares is worth very little? Hell if I know.
So that's when I stopped using the site. I'd thought prediction markets were a straightforward way to profit from having better information/calibration. But then I learned that if you want to profit, this is ultimately a market, and so understanding the fine print and market mechanics (which I don't care about at all) matters at least as much, and probably much more, than having accurate beliefs.
It's particularly upsetting that the site shows you relatively meaningless information like trade volume, while hiding the important information (like liquidity) in an unnoticeable pop-up.
Thanks for your critique and explanation of what you saw! We do need to do a better job at making liquidity more prominent and understandable, as well as making important markets more liquid.
Limit orders should be a much better way of expressing your beliefs without running into the "slippage" problem you're now seeing - but that's not as obvious to use. Curious, did you notice that was an option at all?
This is an important point that trips up a lot of new users. The key thing you have to look at is not just what the current probability (aka price) is, but what it changes to after your bet - what's happening is that as you buy more YES shares the price of the next YES share on offer increases, so if you buy a lot of YES shares at once then the price can go up a lot. You should only make bets where you think the ending probability/price is still in your favor - otherwise, that means your bet is too big and you need to bet less. This also means that if you accumulated a large number of shares, selling them all at once usually won't work well - but what you can do is sell them gradually over time. (This is just like how trading on financial markets work if you are trading large amounts relative to daily market volume.)
Here's an analogy that I think helps explain why large amounts already bet don't really matter: You can think of the 30k already bet as locked in - the market already matched up each previous yes bettor with a no bettor and locked in their bet at specified odds. When you make a new YES bet, you have to be matched with someone else betting NO - the amount that is already locked in between past bettors (whether it's 30k or 0) makes no difference, it doesn't affect how new bets are made; the only thing that matters is how much people are willing to buy NO right now, which is a function of the liquidity pool or limit orders. People buying and selling shares doesn't increase the amount of liquidity, only adding limit orders or adding to the liquidity pool affects that.
I think it's very true that prediction markets emphasize not only making predictions but also trading, and that doesn't interest everyone. I personally came from mostly non-market-based forecasting and think both are great in different ways. I have been advocating for Manifold to support making bets in a more prediction-centric way - many people like the framework of "bet M$100 on YES", but I also want to be able to say "I think YES is about 70% likely, figure out an appropriate bet amount for me". I think there are many different users who find one or the other more natural, so ideally both options can be supported.
Calling Kalshi the monopoly is kind of rich. PredictIt absolutely had monopolistic pricing and a weird no-action letter no one else could compete with. Kalshi has a first mover advantage but not a monopoly. Anyone can — and will — go through actual regulatory channels to make competitive markets. Markets are a pretty slam dunk case for regulatory rule setting and oversight as well.
Also VPNs are cute, but the issue with non-US market use for US citizens *I*s *R*eally *S*omething else entirely….
Is it illegal to bet on them, or merely illegal to offer them*? If the latter, there's no issue - you can tell the IRS the truth.
*Like how e.g. downloading pirated stuff isn't illegal, but uploading it is. Or how it's illegal to operate a pharmacy without a licence, but if you buy stuff from an illegal pharmacy that's often not illegal (not always, though).
Well even if it isn't illegal, you owe US taxes on your international income. But the site is presumably reporting it to local tax authorities, so you could end up having your profits taxed twice, and/or having to explain to two sets of tax authorities why your income isn't taxable by them. It's painful enough with perfectly legal stock dividends that have tax treaties, doing it with a prediction site feels like a no-win situation to me.
And then of course as prediction markets grow they'll need real liquidity. Try telling the Chief Risk Officer of a reinsurance firm that you can buy stuff from illegal pharmacies legally sometimes (but not always) and see how quickly he says no,
Someone should lobby to transfer jurisdiction over prediction markets to a new regulatory agency.
Removing CFTC jurisdiction would suffice.
I think the CSPI competition wasn't setup correctly for reasons well described in this comment from their announcement: https://www.cspicenter.com/p/introducing-the-salemcspi-forecasting/comment/8235817
///
This doesn't work as well as you think - financial markets tend to break when participants are optimizing for the probability of finishing in the top N, especially when N is small. If I were trying to maximize my probability of winning this tournament in a blended pool of rational and irrational agents, I would be:
1) taking way more risk than I would if my goal were simply to maximize my portfolio's returns
2) taking risks that give me different risk exposures from other participants, even if the arithmetic expectation isn't in my favor.
For example, if the true probability of an event happening is 90%, and various market participants bid it up from 50% to 88%, well, I am probably more likely to sell it than buy it at those prices. The intuition is that if the event resolves YES, I will have bought it at the worst price among all the people who bought YES, but if the event resolves NO then I will have sold it at the best price among the people who sold YES.
Requiring people to trade a variety of markets also doesn't do much - it's pretty easy for people to buy YES at 99% on a variety of already-basically-resolved markets to meet this requirement, especially when those markets will formally resolve quickly.
All in all I expect that this competition tests the skill of "how good are you at gaming formal systems" a lot more than it tests the skill of "how good are you at making well calibrated predictions about the world".
Source: I was a quantitative trader for 2 years and they made us play a bunch of games market trading games with solutions like these during training.
///
In the updated rules post, they say they'll subjectively evaluate betting behavior to decide the winners instead of just looking at profits in order to discourage this.
> I love Manifold as a company but I’m sometimes skeptical of what I see there - there are too many really low-liquidity markets, too many sources of extra play money, too many stupid “the probability on this market will be under 49% on August 31” troll questions, too little desperation keeping people on their toes. The Salem subtournament solves all those problems
I agree that most of those are problems. But it's worth calling out this one:
> too many stupid “the probability on this market will be under 49% on August 31” troll questions
Questions in that genre aren't problems. They're valuable sources of empirical insight into the general area of "Keynesian beauty contests", which are of considerable interest.
"You know a prediction market has really made it when its insane questions about Donald Trump start getting close to zero. Insight has almost made it."
Scott, as one of the internet's premier prediction market evangelists, you should be better than this. Even the market "This market resolves to true at the end of 2022" wouldn't get much lower than 2%. due to the risk-free rate.
This was all new to me, but I found it extremely informative and fascinating. Love your subliminal suggestions on how to use the foreign exchanges!
PredictIt's New Zealand arm was already shut down years ago when some local anti-money laundering legislation was passed. I've disliked anti-money laundering laws as a concept ever since.
Which law or laws allow Americans to gamble at a casino in e.g. Macao but not, while in e.g. Macao, bet real money on a non-US based website?
Hey Scott, I've been following and reading your blog for sometime, and I’m fairly interested in prediction markets as well! I am hopeful that regulatory issues on the US are solved and that the projects you mentioned can grow, not just those that work with fake money.
Have you heard about Polkamarkets? I don't recall you ever mentioning them. They are decentralized and autonomous, but the feature I like the most is that any person can create prediction markets on real-life events.
You can also make forecasts, and resolve markets. They’re built on multichain EVM, Moonriver and Moonbeam networks atm.
Not a very big community, they lack liquidity and volume atm but their app has an incredibly ux/ui, it’s very functional tbh. They’re using DeFI and NFTs in an interesting and entertaining way too
"Nameless corporate composers"?
Max Martin ( https://en.wikipedia.org/wiki/Max_Martin ) isn't a household name the way the performers he wrote and produced music for are, but the only songwriters that have had more #1 Billboard hits than he did were Paul McCartney and John Lennon, who are kind of a special case...
The 10th avatar of Vishnu is Kalki, not Kalshi: https://en.wikipedia.org/wiki/Kalki. Kalshi is an Arabic word meaning everything: https://www.forbes.com/sites/frederickdaso/2019/04/03/kalshi-an-mit-betting-startup-is-allowing-people-to-bet-on-anything/
You should use "it" as a pronoun for organizations rather than "they". Otherwise there's an inconsistency in verb conjugation: "CFTC is" but "they are". I know it's currently in vogue to refer to organizations in plural but it's grammatically incorrect and leads to such jarring inconsistencies.
Actually Predicit’s violations are more serious than I realized. The initially terms are for “contracts consisting of two sub markets regarding binary election outcomes and economic indicators”
Most of predictit’s contracts do not comply.
Smarkets.com is another non-US option.
> I think Metaculus’ defense would be that their aggregation method lets them weight the guesses of people with good track records more highly
I should have been clearer about this in my post, but I was mostly taking aim at what Metaculus calls the "Community Prediction" -- which is just a (recency-weighted) average of all the predictions, like a literal poll. (As opposed to the "Metaculus Prediction," which is weighted based on track record.)
The reason I focused so much on the Community Prediction is that, when people talk about Metaculus forecasts, they're almost always talking about the Community Prediction. For example, everyone talking about moves in Metaculus graphs is talking about the Community Prediction, since no one can even see the Metaculus Prediction until a question has closed (unless they spend "Tachyons" to get a sneak preview for a particular question).
I agree that doing track-record-weighted aggregation would resolve some of my concerns. But if you ask the question "does Metaculus do track-record-weighted aggregation?", the answer is almost always "not for the numbers we're discussing, no."