Just to state the obvious: ACX's homepage design has just changed, instead of showing thumbnail and titles for every post, the first 1-2000 words of each post are shown in-full. (One implication is that you do not see the thumbnail images on new posts, which is a little sad.)
I sort of assumed this was at Scott's request, since it makes the layout much more like the WordPress experience on SSC. As far as I know he only adopted thumbnails because Substack's default style makes them so prominent.
I preferred the thumbnails, because it meant much less scrolling to get to the slightly-old posts whose comment threads I was still following while only adding one click to the "difficulty" of reading the very latest post.
If this isn't going to change back, how about a sidebar with the archive in it, even though that means that the site wastes less of my desktop screen with empty white space (oh no!), and actually looks different on mobile vs. desktop (the horror!).
At least the direct link to the archive still works. *edits bookmark*
I'm having some kind of weird bug where the Polymarket post stays on top of the other format no matter how many other posts I make. This is my hack to get around that. Substack tech support is looking into it; once they solve it, I'll change back.
Thanks for the corrections and updates. Good to know Polymarket is still active with both US and international options. The prediction contest deadline extension to next week is helpful too.
For those interested in exploring alternative gaming platforms, you might check out https://nullbrawlpro.com/de/, it's a private server for Brawl Stars with unlocked features and resources. Seems to offer a different take on mobile gaming.
The collaboration idea between ACX predictions and prediction markets sounds interesting. Hope you get some creative proposals through the feedback form before the deadline.
That clarification about Polymarket makes sense. I also like the idea of comparing reader predictions with prediction markets, especially if the data can reveal where crowd forecasts and market expectations diverge. There could be some really interesting findings there. The collaboration idea is worth exploring too, since having access to the data before publication could open up some creative experiments.
It’s interesting how comparing different sets of data can sometimes reveal patterns you wouldn’t notice otherwise. Even something outside the research space, like tracking the progress and results of a fence cleaning project like https://powerwashinglv.us/fence-power-washing/, can give you useful before-and-after data to compare.
Portfolio construction with an ensemble of expert forecasts vs. a real-money market is harder than it looks. (I left a comment through the form.)
To figure out if this is to good to be true, I just consulted my fraud detection rock. It read "IT'S TO GOOD TO BE TRUE".
Also, $17632 is not even close to enough to pay for a lesswrong.substack.com subscription, which I just discovered was a thing, in 2021, apparently.
Just to state the obvious: ACX's homepage design has just changed, instead of showing thumbnail and titles for every post, the first 1-2000 words of each post are shown in-full. (One implication is that you do not see the thumbnail images on new posts, which is a little sad.)
I sort of assumed this was at Scott's request, since it makes the layout much more like the WordPress experience on SSC. As far as I know he only adopted thumbnails because Substack's default style makes them so prominent.
I preferred the thumbnails, because it meant much less scrolling to get to the slightly-old posts whose comment threads I was still following while only adding one click to the "difficulty" of reading the very latest post.
I would appreciate a reversion to the previous appearance, too.
I agree.
If this isn't going to change back, how about a sidebar with the archive in it, even though that means that the site wastes less of my desktop screen with empty white space (oh no!), and actually looks different on mobile vs. desktop (the horror!).
At least the direct link to the archive still works. *edits bookmark*
What I'd really like is an index like the old SSC had, one item per line (not the front page thing Freddie deBoer has, either).
I'm having some kind of weird bug where the Polymarket post stays on top of the other format no matter how many other posts I make. This is my hack to get around that. Substack tech support is looking into it; once they solve it, I'll change back.
..essentially, all models are wrong, but some are useful
-George Box
The new ACX front page doesn't seem to have a blogroll anymore or a link to DSL. Is it just me?
Solid point! I rly like(d) the blogroll. Including the latest additions like Dynomight and that other one. (I don't know what DSL is.)
Thanks for the corrections and updates. Good to know Polymarket is still active with both US and international options. The prediction contest deadline extension to next week is helpful too.
For those interested in exploring alternative gaming platforms, you might check out https://nullbrawlpro.com/de/, it's a private server for Brawl Stars with unlocked features and resources. Seems to offer a different take on mobile gaming.
The collaboration idea between ACX predictions and prediction markets sounds interesting. Hope you get some creative proposals through the feedback form before the deadline.
That clarification about Polymarket makes sense. I also like the idea of comparing reader predictions with prediction markets, especially if the data can reveal where crowd forecasts and market expectations diverge. There could be some really interesting findings there. The collaboration idea is worth exploring too, since having access to the data before publication could open up some creative experiments.
It’s interesting how comparing different sets of data can sometimes reveal patterns you wouldn’t notice otherwise. Even something outside the research space, like tracking the progress and results of a fence cleaning project like https://powerwashinglv.us/fence-power-washing/, can give you useful before-and-after data to compare.