Click here to begin voting. Click here for a general overview of the 2020 Review.
We're now in the final week of the LessWrong 2020 Review. We've had over a month of preliminary voting and discussion. Now it's time to finalize our votes. In this post I cover how to vote, why to vote, and what will happen with the outcome (in no strict order).
The winning posts will be assembled into the longterm annals of the Best of LessWrong. This year, each top-rated post will be displayed with a donation button for its author, so readers can directly show their support for posts that were important to them.
The LessWrong Review has two major goals.
Voting sends a signal about which posts were most important. This can feed into various compilation and distillation efforts.
All users registered before 2020 can vote. The LessWrong curation team will weight the votes of users with 1000+ karma more highly when assembling sequences, books, or other projects.
Go to the Review Voting page. There, you'll see posts that you haven't yet voted on sorted first, with posts you've previously given a karma-vote to sorted to the very top (they have a green stripe along the left side. Strong upvotes have a darker green).
The posts will include the reviews that got written about that post. This is intended to help you make an informed vote. You can read reviews that look interesting or highly upvoted to get a better sense of how the post held up.
You vote by clicking buttons that assign a post a score. A score of 1 means roughly "this post was good." A score of 4 means "this post was quite important". A score of 9 means "this post was extremely important." A vote of 0 means "I don't have a strong opinion."
(A vote of "4" is 4x as strong as a vote of "1", a vote of "9" is 9x)
If you intuitively mark posts as "good/important/extremely-important", you'll probably do fine expressing your votes.
Okay but I'm a nerd and like knowing the details of things. How does this work under the hood?
The system is built on top of a quadratic voting system. In quadratic voting, you have a limited number of points to spend. You can vote things more strongly, but the cost of a vote increases at a quadratic rate. So:
You have 500 points to spend. If you spend more than 500 points, your votes get proportionately downweighted, which looks like this:
It's up to you whether to vote a lot of things at the strongest setting (and have the votes count less), or to limit yourself so that your strongest votes count for a full 9 points.
I'm excited to see how various posts fair. I'm personally excited for people to take the time reflecting a bit, seeing how various posts compared against each other and against an overall standard of "what was important?". Use your own judgment about what that means, and help shape the longterm record of LessWrong history. :)
If you have any questions, bug reports, mild irritations or suggestions on UI improvements, let me know!
We're still making the final call of whether/how to make the posts into physical books. We're considering a lot of different options, including "similar books to last year", "custom textbooks that are more 'inspired by' than directly taking the winning posts", and "no books."
The quadratic formula here is (n^2 + n)/2, i.e. if n is 4 (a "medium vote"), you get a cost of 10. When you overspend, it divides the cost of all your votes proportionately by x/500, where x is however many points you spent. i.e. if you spent 550 points it'd divide all your vote costs by 1.1. Then, uses the inverse of the quadratic formula to calculate new vote-values.
(n^2 + n)/2
Thank you lightcone team for continuing to make these happen. I expect I owe several of you a free gratitude coffee.
Unexpected behavior / minor bug: Some posts from 2020 have the "Vote on this post for the 2020 Review" section, despite not being eligible for Final Voting because they lack a review. Here's a random example.
Feedback on the quadrating voting implementation: This might have already been mentioned by others, but I wish there were no -1 option, as this seems to incentivize weird voting behavior like downvoting all essays you haven't upvoted, rather than upgrading your +1 votes to +4.
I haven't done the full math, but I think this problem would still exist somewhat if the cheapest negative option was -2 (which would cost 3 points), but not if it was was -3 (which costs 6 points). There are 122 reviewed posts this year, and you have 500 points. So downvoting almost all of them to -1 would only cost ~122 points, leaving you 378 or ~75% of your points for upvotes; whereas downvoting to -2 would cost ~366 points and leave you 134 or ~25% of your points for upvotes.
We're still making the final call of whether/how to make the posts into physical books. We're considering a lot of different options, from "
The footnote cuts off here.
Oh, I mistakenly translated "sort by your vote" into "sort by Needs Vote." Your right, sort by "your vote" is just broken right now. (good news is that "sort by Needs Vote" is approximately the same thing, except you have to skim past things you didn't vote on yet)
I'm working on a fix now.
The hover texts for the -9 to +9 votes are inconsistent:
Intended behavior by "sort by your vote" is to first show you posts you haven't voted on (with "posts you previously karma voted on" clustered to the top") followed by posts that you have voted on (in descending order of vote-strength).
Does your experience match that description? (admittedly that's not super legible). If not, there might be some bugs in addition to not-being-super-legible.
Oooh, since the 0-strength vote is automatically greyed out to highlight it as the default, I didn't realize voting 0 would do anything, so I never did. But it appears that the Sort features do take that into account.