Will this paper be widely agreed to have shown that Othello is solved?
8
95
แน€70
May 6
91%
chance

This preprint claims to show that Othello is solved: https://arxiv.org/abs/2310.19387

This question resolves positively if, in my opinion, it has been accepted as correct in 6 months.

By accepted as correct, I mean that some scrutiny by credible individuals has been applied, and the claim of the paper is accepted as correct. This may involve the author of the paper providing more code or data.

If the paper goes effectively unnoticed for 6 months, this resolves as N/A.

Get แน€200 play money
Sort by:
bought แน€41 of YES

Gave it a quick skim. Did not read the source code or verify any of that. I have some relevant interested-amateur-level knowledge here; I've contributed to Go programs, and written my own programs that play Go, Checkers, and Amazons. I understand alpha-beta search and the optimizations they used.

As best I can tell they didn't do anything especially fancy with room for mistakes of theory or math; they just took the obvious approach with some very careful optimization. I think this means that if they got the clever pieces wrong, it would mostly have impacted the efficiency and not the end result. Building off a well-respected program is a strong start.

Assuming the authors are not lying (why would they be?), I give this a very high chance of being accepted.

@EvanDaniel From a brief look I couldn't tell if they adequately justified their claim that solving the subset of positions they did actually proves that they solve the game.