Human Chess is a chess variant where playing the top engine move is forbidden • Hacker News Board

Re: Human Chess is a chess variant where playing the top engine move is forbidden

#241

Post by brandnewaccount » Fri, Dec 02, 2022, 8:20 PM UTC

jl6 wrote: post #53

mattmaroon wrote: post #32

Earlier quoted context omitted.

Nah, you’d spot one pretty fast when you blundered and they went to take advantage of it. Instead you’d more likely have the opposite problem where you’d veto after a blunder but still be at such a huge disadvantage that it wouldn’t matter much. It would be pretty neat between players of similar skill level though, then I could see the hoarding taking place.

True, and I can see some fun mind-games where a player might try baiting an opponent into wasting their veto on an apparently-strong move, or by intentionally playing a weaker move that still somehow looks strong but actually masks a now-unvetoable killer move…

Maybe for players under say 1800 elo online, but for players above that this won't work -- "bluffing" isn't really a thing until you're at the very very highest levels of chess, and even then the bluffs are only during the openings and if they call your bluff you are only worse by 0.1-0.4 at the most.

Re: Human Chess is a chess variant where playing the top engine move is forbidden

#242

Post by yoyo348 » Sun, Dec 04, 2022, 2:49 AM UTC

Nice variant

Re: Human Chess is a chess variant where playing the top engine move is forbidden

#243

Post by nimih » Mon, Dec 05, 2022, 12:46 AM UTC

ratorx wrote: post #235

nimih wrote: post #232

Earlier quoted context omitted.

With this variant, since there are a finite number of legal board positions in chess, and Human^n chess is simply chess with a finite number of moves prohibited at each board position (claiming is the same as resigning, in this formulation), there are a finite number of distinct Human^n chess, so the mapping Human^n-1 chess -> Human^n chess must eventually reach a cycle (potentially of length 1, i.e. a fixed point).

I’m not sure about a straight-up cycle. I think it requires determinism, whereas the engine is not necessarily deterministic (eg. NN based ones like AlphaZero). I think it’s possible that at least 1 position (and probably a lot more) will have more than 1 optimal solution. If the engine is not fully deterministic, then it’s possible instead of a normal cycle, there are a fixed set of strategies at each n that form a…

That's a good point about non-determinism, although I wonder if there are known convergence/stability results in the ML literature that allow you to effectively ignore that detail (i.e. it seems plausible that you could get something like "human^n chess always stabilizes at some fixed amount of training time/computational power"). You can also just fix the randomization seed, but that's obviously a less satisfying result.

> The other thing that would compound this is that no current chess engine solves the game fully. There would be even more positions that have multiple “optimal” solutions if the engine only looks ahead to bounded x.

I'm not sure this is an obstacle; we're explicitly excluding the computer's preferred move, rather than the necessarily optimal move, after all. You could easily play human chess with the engine from (e.g.) Battle Chess, which honestly is sort of an interesting idea in its own right.