Gukesh becomes the youngest chess world champion in history

Post by renewiltord » Fri, Dec 13, 2024, 4:05 AM UTC

svat wrote: post #379

le-mark wrote: post #362

Earlier quoted context omitted.

The winning move is at 4:01:45 for anyone looking (like I was).

Just curious: The comment you're replying to had the link with timestamp 4:01:38 which is basically just before the move happens; is that not enough as it is?

Post by lmm » Fri, Dec 13, 2024, 4:25 AM UTC

sheepscreek wrote: post #360

School-Cotton wrote: post #298

Earlier quoted context omitted.

It's not hard to make a chess bot that plays at a 1300 strength, i.e. its rating would converge to 1300 if it were allowed to compete. But it will not play like a 1300-rated human. It would play like a superhuman genius on most moves and then make beginner-level blunders at random moments. Making one that realistically plays like a human is an unsolved problem.

I think this can be achieved with some ease with a machine learning model. You will have to train it on games between 1300-rated players and below. A transformer model might work even better in terms of the evenness of play (behaving like a 1300 rated player throughout the game).

Post by lxgr » Fri, Dec 13, 2024, 5:00 AM UTC

paulddraper wrote: post #251

tomatovole wrote: post #164

Earlier quoted context omitted.

Is there a metric I can look at in engine evaluations to determine when a situation is "risky" for white or black (e.g., the situation above) even if it looks equal with perfect play? I've always been interested in understanding situations where this is the case (and the opposite, where the engine favours one side but it seems to require a long, hard-to-find sequence of moves. Playing out the top lines helps if equal…

You can evaluate on lower depth/time. But even that isn't a good proxy. Humans cannot out-FLOP a computer, so they need to use patterns (like an LLM). To get the human perspective, the engine would need to something similar.

Post by hilux » Fri, Dec 13, 2024, 5:30 AM UTC

eitally wrote: post #208

cenamus wrote: post #148

Earlier quoted context omitted.

A lot of exceptions sadly. But with the amount of neuroticism you see with top athletes that's to be expected I suppose

I disagree. I think what you're saying are "a lot of exceptions" are primarily going to be in what are historically referred to as "revenue sports" in the US -- football & basketball -- and also in individual sports where personal marketing is key to financial success (e.g. sponsorships). I don't see it as unexpected for there to be big egos and boisterous personalities in sports where individuals are hugely rewarded…

Post by hilux » Fri, Dec 13, 2024, 5:33 AM UTC

auc wrote: post #373

stouset wrote: post #352

Earlier quoted context omitted.

> Personally, I'm on the side which thinks that this format is a total stagnation. I was with you until this WCC. In almost every game, Gukesh took Ding out of prep extremely early. It wasn't always a success for him either! Leaving the opening book typically means you accept an objectively-worse position, but one that your opponent has to spend a significant amount of time finding the right ideas in. Even in the cas…

This lacks some context. Gukesh took Ding out of prep bc Gukesh prepped for 7 months, while Ding prepped for 3 weeks and basically wanted to retire from chess. Ding would be out of prep by move 8 as white and be an hour down on time very consistently.