Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Stockfish 14

stockfishchess.org

Post reply on HN
Print view View original

71–80 of 207 posts

Re: Stockfish 14

#71

Post by keerthiko »

SamBam wrote: post #52
tus89 wrote: post #25

Is it just me or is battle-of-the-AI chess more exciting than humans?

Minor thought, but one of the things I love about high level AI chess, like Stockfish vs AlphaZero, is seeing how their ratings of positions change over the course of the game. I had realized a while ago that, as a human, the computer will give your position a score, and then you make a move, and your score can pretty much only stay the same (if you make a "perfect" move) or go down. Much of the time, it goes down. T…

> Two humans playing each other, it's just a question: who's score goes down less each time they make a move? It became a little sad. It seemed like either you can make the right move, or you make a sub-optimal move, and the winner is simply the one who makes the fewer sub-optimal moves.

What you describe is when a perfect chess-playing computer (approx AlphaZero) observes two humans playing. What humans observe watching two humans playing (including the participating humans) is very similar to what you described AlphaZero vs Stockfish as. The only difference is we don't ask human players to ascribe a score to their opponent's move (and wouldn't expect it to be accurate)

Re: Stockfish 14

#72

Post by astrange »

ivalm wrote: post #9
cellis wrote: post #6

Earlier quoted context omitted.

Isn't the difference though, that Leela is a pure RL implementation whereas NNUE/SF still uses some tablebase stuff?

Leela uses mcts, sf does not. SF also trains on human positions, while I think Leela is self play only.

AlphaZero and co only trained via self-play because that was their research goal, and they looked unwinnable because they were the first to get this kind of neural net working, but it doesn't seem like it's the best option.

AZ notably got completely tilted once it started losing, doesn't necessarily recognize strange positions you can't normally get into, and doesn't care about its win margin at all.

Re: Stockfish 14

#73

Post by zucker42 »

jaimex2 wrote: post #16

Rematch Alpha Zero. Have power consumption restraints. Go!

Stockfish would beat AlphaZero handily. Even the original paper's test conditions were not fair, so Stockfish may have been much better than the results of the original matches made it seem.

Re: Stockfish 14

#74

Post by zucker42 »

bigpumpkin wrote: post #26
jaimex2 wrote: post #16

Rematch Alpha Zero. Have power consumption restraints. Go!

Good idea. or have a dollar amount limit on the hardware

Given that AlphaZero will presumably never be publicly available, I think you might be interested in TCEC which has fair fights between Stockfish and LeelaChessZero (which Stockfish has won recently).

Re: Stockfish 14

#75

Post by vikingerik »

SamBam wrote: post #52
tus89 wrote: post #25

Is it just me or is battle-of-the-AI chess more exciting than humans?

Minor thought, but one of the things I love about high level AI chess, like Stockfish vs AlphaZero, is seeing how their ratings of positions change over the course of the game. I had realized a while ago that, as a human, the computer will give your position a score, and then you make a move, and your score can pretty much only stay the same (if you make a "perfect" move) or go down. Much of the time, it goes down. T…

In a game of perfect information and no randomness, there are ultimately only two kinds of moves: those that preserve your current best forcible outcome (win or draw in chess*), and those that blunder that into a worse result given continued perfect play by the opponent.

Everything else like a positional score or centipawns or even classic material points is an abstraction, that we use to summarize because we don't have unbounded or sufficient computing power to solve all possible continuations. That score apparently going down is only an artifact of our limited ability to evaluate it; the only real scores are 0/½/1 for lose/draw/win. If you make mistakes, your score will evetually drop by those quantizations; we just typically don't know exactly when, except in endgame situations pared down enough to be computationally tractable.

And it's impossible to raise your estimated score, because that estimation assumes you continue to play perfectly. There's no such concept as a better-than-perfect move to raise your expectation over what was already calculated, since that calculation already includes all your best possible moves.

* (Other gradiations between win/lose/draw are possible in such a game. Chess doesn't have such, but imagine playing Go for a dollar per point, where nuances smaller than swinging a win or draw still matter.)

Re: Stockfish 14

#76

Post by amenghra »

stouset wrote: post #58
SamBam wrote: post #52

Earlier quoted context omitted.

Minor thought, but one of the things I love about high level AI chess, like Stockfish vs AlphaZero, is seeing how their ratings of positions change over the course of the game. I had realized a while ago that, as a human, the computer will give your position a score, and then you make a move, and your score can pretty much only stay the same (if you make a "perfect" move) or go down. Much of the time, it goes down. T…

> But when AlphaZero plays, and you watch Stockfish's score, the reason it wins is that it makes moves Stockfish thinks is poor, so it rates AlphaZero's moves poorly, and then all of a sudden it has an oh shit! moment when it realises that AlphaZero is actually ahead, and its score jumps. It's really a look inside the computer's head while it's being beaten by a better player. That's more or less a description of wha…

Right. Think about all the times you would answer “no” to “would you like to switch sides with your opponent right now?”, even as you are about to lose a few moves later.

Re: Stockfish 14

#77

Post by dangero »

anikan_vader wrote: post #63
warent wrote: post #54

Earlier quoted context omitted.

Nitpick, certainly not thousands. When you're first starting out even if your elo is like 400, you'll be lucky to break 1000 after a year unless you study very diligently. The gap from 1000-2000 then takes multiple years (probably like 4-5 of diligent study) Climbing chess elo is very hard and slow!

I think reaching 1400 elo after a year of decent study after learning the rules is not too unreasonable. To reach 1400 in a year you would need to be either talented or hard-working. Climbing the elo ladder does get much harder though, as you mentioned. To reach 2000 in two years you would need to be both hard-working and talented. To reach 2850 in a lifetime you have to be hard-working, talented, and a particular No…

In which format? I highly doubt that someone could do that in bullet unless they played full time.

Bullet players are already on average a higher percentile of chess players so the elo is skewed

Re: Stockfish 14

#78

Post by awb »

elias94 wrote: post #49
jeremyjh wrote: post #33

And chess.com still uses version 8...

True, but who uses chess.com anymore?

They’re a great site for chess and chess variants. And their live commentary of chess events is actually exceptional.

Re: Stockfish 14

#79

Post by kohlerm »

adgjlsfhk1 wrote: post #62
ivalm wrote: post #9

Earlier quoted context omitted.

Leela uses mcts, sf does not. SF also trains on human positions, while I think Leela is self play only.

SF14's net was actually trained on lc0 data. Human data is way too weak for SF to learn from.

Yes and I think it's really generous of the LC0 to "donate" that data. On the other side LC0 uses SF during the training process since a while.

Re: Stockfish 14

#80

Post by adgjlsfhk1 »

kohlerm wrote: post #79
adgjlsfhk1 wrote: post #62

Earlier quoted context omitted.

SF14's net was actually trained on lc0 data. Human data is way too weak for SF to learn from.

Yes and I think it's really generous of the LC0 to "donate" that data. On the other side LC0 uses SF during the training process since a while.

It makes sense. They're both open source programs working towards better understanding of chess. There are actually a few people who develop for both. They are obviously very different types of engines, but they are much closer to friendly rivals than enemies.
Post reply on HN