I've created 14 markets asking, "What is Manifold's favorite X?" that resolve based on a survey where participants can rank each work as "Excellent", "Good", "Okay", "Bad", or "Haven't seen/read/played/heard". The way I will choose Manifold's favorite is just by taking 2*(no. of "Excellent" rankings) + (no. of "Good" rankings) - (no. of "Bad" rankings). This means that a work that more people are familiar with has an advantage, since it's more likely to get more Excellent or Good rankings.
Although I think this is the most faithful interpretation of "favorite", it's also interesting to see which work Manifold judges as the best, that is, "What do the people who have actually seen/read/played/heard the work think of it?" To determine that, I could exclude all the "Haven't seen/read/played/heard" ratings, and take the average, rather than the sum, of the remaining ratings to see what the average person who is familiar with the work thinks. For how many of my markets will this produce a different result than the "favorite"?
If there's a tie when using any method, I'll count it as a full change unless both methods result in the exact same options being tied (in other words, I'm not bothering with "partial agreement" like I did in the dwarf planet market).
The Manifold's favorites markets are:
/JosephNoonan/what-is-manifolds-favorite-movie
/JosephNoonan/what-is-manifolds-favorite-book
/JosephNoonan/what-is-manifolds-favorite-tv-show
/JosephNoonan/what-is-manifolds-favorite-print-co
/JosephNoonan/what-is-manifolds-favorite-webcomic
/JosephNoonan/what-is-manifolds-favorite-song
/JosephNoonan/what-is-manifolds-favorite-video-ga
/JosephNoonan/what-is-manifolds-favorite-board-ga
/JosephNoonan/what-is-manifolds-favorite-roleplay
/JosephNoonan/what-is-manifolds-favorite-short-st
/JosephNoonan/what-is-manifolds-favorite-poem
/JosephNoonan/what-is-manifolds-favorite-podcast
/JosephNoonan/what-is-manifolds-favorite-youtube
/JosephNoonan/what-is-manifolds-favorite-play
If I add any others, they won't count for this market.
See Plasma's Manifold Survey for other questions about the survey.
The survey is officially out! You can take it here: https://forms.gle/xZqWVxuY5irgLigu9
Related questions
I created a version of this market that avoids the "unknown lunatic wins" scenario that can happen in average-based voting:
The survey is officially out! You can take it here: https://forms.gle/xZqWVxuY5irgLigu9
Going to post the same thing I said on Discord here:
Interested to see the results of this one. My intuition is that a lot of categories will have at least one work that's a hidden gem, cult classic, or a work with a small but dedicated fanbase, and that these works will tend to get higher average ratings from those who know them, but won't win the position of favorite because not that many people do. Plus, a lot of the categories have a ton of works, which just makes it more likely that there will be a split between different methods of judging them, especially if it is close. And then of course, there's also the possibility that someone will submit a work that they think is excellent, but no one else has even heard of, causing its average rating to be "Excellent".
So basically, I think there's a good chance that the number will be higher than the current prediction.
