Certain top engines are only good at playing strong chess and that is all ?
Moderators: hgm, Rebel, chrisw
-
- Posts: 5582
- Joined: Wed Sep 05, 2018 2:16 am
- Location: Moving
- Full name: Jorge Picado
Certain top engines are only good at playing strong chess and that is all ?
I suspect that certain top engines like Stockfish are very strong at playing normal or standard chess or Chess960, but as far as we know, that is the limit; but I would like to find out if given Odds like a Knight Plus a pawn if Stockfish can beat my friend rated 2458, tomorrow we will find out at 15'+10". What is your prediction can stockfish also play well giving Odds
Do NOT worry and be happy, we all live a short life
-
- Posts: 12540
- Joined: Wed Mar 08, 2006 8:57 pm
- Location: Redmond, WA USA
Re: Certain top engines are only good at playing strong chess and that is all ?
The statement is vague. If Stockfish is running on an iphone, I would take the GM player + knight + pawn.
If Stockfish is playing on Bojung Guo's strongest machine, I choose Stockfish.
I don't have enough data to choose anything in-between. I guess that if the GM plays anticomputer chess style, he can at least get draws.
If Stockfish is playing on Bojung Guo's strongest machine, I choose Stockfish.
I don't have enough data to choose anything in-between. I guess that if the GM plays anticomputer chess style, he can at least get draws.
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.
-
- Posts: 4556
- Joined: Tue Jul 03, 2007 4:30 am
Re: Certain top engines are only good at playing strong chess and that is all ?
That's not chess.Chessqueen wrote: ↑Fri May 29, 2020 10:49 pm I would like to find out if given Odds like a Knight Plus a pawn if Stockfish can beat my friend rated 2458
Your "2458" friend hasn't been rated at Odds like a Knight Plus a pawn chess, so his rating for this is meaningless.
-
- Posts: 5960
- Joined: Sun Jan 10, 2010 6:15 am
- Location: Maryland USA
Re: Certain top engines are only good at playing strong chess and that is all ?
I don't agree that it has much to do with hardware, at least assuming a modern pc, not a phone. The problem is that Stockfish is not optimized to play well when way down in material, it will play to postpone the loss, which is very different from maximizing chances. I think that any GM or strong IM will generally win and rarely lose to Stockfish at knight odds, never mind knight plus pawn, at any normal rapid time control, even if it has 1000 cores. That's not to say that a version of Stockfish couldn't be created that would win at knight odds, but someone has to do the work. The only engine that might beat him at knight odds, aside from possibly Komodo MCTS, is the old Lc0 11248 (or 11258) network on a 2080 GPU. I came very close to beating Stockfish 11 myself at two pawn odds (c2 and f2) at 10' + 5" yesterday, it barely saved a draw in a clearly lost rook endgame. That's probably less than half of knight odds.Dann Corbit wrote: ↑Fri May 29, 2020 11:48 pm The statement is vague. If Stockfish is running on an iphone, I would take the GM player + knight + pawn.
If Stockfish is playing on Bojung Guo's strongest machine, I choose Stockfish.
I don't have enough data to choose anything in-between. I guess that if the GM plays anticomputer chess style, he can at least get draws.
Komodo rules!
-
- Posts: 3657
- Joined: Wed Nov 18, 2015 11:41 am
- Location: hungary
Re: Certain top engines are only good at playing strong chess and that is all ?
If Stockfish plays with odds it is equal to play with big negative contempt.
Can play Stockfish with negative contempt? Yes, he can.
So? What is the problem?
Can play Stockfish with negative contempt? Yes, he can.
So? What is the problem?
-
- Posts: 5563
- Joined: Tue Feb 28, 2012 11:56 pm
Re: Certain top engines are only good at playing strong chess and that is all ?
How is playing with big negative contempt equal to playing with one knight less??
-
- Posts: 3657
- Joined: Wed Nov 18, 2015 11:41 am
- Location: hungary
Re: Certain top engines are only good at playing strong chess and that is all ?
Stockfish strives to reduce the negative contempt for reaching near zero position value if he can do it.
Obviously in the case of a figure disadvantage he can lessen the drawback if he can find a tactical opportunity.
I would be curious to know with a knight disadvantage the Stockfish is the better or the Komodo.
I think Stockfish is better near the same Elo as it is the difference between Stockfish and Komodo.
In the case of a Bishop disadvantage the situation is different because of the pair of Bishop.
I propose to Mr. Kaufman to run a match between Stockfish and Konodo with a knight odd.
I am also very curious to see the result even if I am not a believer of this type of matches.
-
- Posts: 5960
- Joined: Sun Jan 10, 2010 6:15 am
- Location: Maryland USA
Re: Certain top engines are only good at playing strong chess and that is all ?
Engines can only give knight odds to other engines when the rating difference is something like 800 elo (depending on time limit, threads, etc.). I did some tests a while ago in which I had both Komodo and Stockfish give knight odds to a much weaker engine (it might have been a crippled Komodo level, I forget the details) a while ago, one weak enough (like about Magnus Carlsen level!) to make for fairly close matches, and I recall the results of Stockfish and Komodo were very similar, but I don't remember enough details. I could redo the experiment at some point against some suitable engine, with high Contempt settings.corres wrote: ↑Sat May 30, 2020 7:07 pmStockfish strives to reduce the negative contempt for reaching near zero position value if he can do it.
Obviously in the case of a figure disadvantage he can lessen the drawback if he can find a tactical opportunity.
I would be curious to know with a knight disadvantage the Stockfish is the better or the Komodo.
I think Stockfish is better near the same Elo as it is the difference between Stockfish and Komodo.
In the case of a Bishop disadvantage the situation is different because of the pair of Bishop.
I propose to Mr. Kaufman to run a match between Stockfish and Konodo with a knight odd.
I am also very curious to see the result even if I am not a believer of this type of matches.
Komodo rules!
-
- Posts: 3657
- Joined: Wed Nov 18, 2015 11:41 am
- Location: hungary
Re: Certain top engines are only good at playing strong chess and that is all ?
I think it is only sense of a Stockfish-Komodo match, from at least 2 x 50 opening positions with alternated colors plus alternated odd, that is 4 x 50 = 200 games (at least).lkaufman wrote: ↑Sat May 30, 2020 8:03 pmEngines can only give knight odds to other engines when the rating difference is something like 800 elo (depending on time limit, threads, etc.). I did some tests a while ago in which I had both Komodo and Stockfish give knight odds to a much weaker engine (it might have been a crippled Komodo level, I forget the details) a while ago, one weak enough (like about Magnus Carlsen level!) to make for fairly close matches, and I recall the results of Stockfish and Komodo were very similar, but I don't remember enough details. I could redo the experiment at some point against some suitable engine, with high Contempt settings.corres wrote: ↑Sat May 30, 2020 7:07 pmStockfish strives to reduce the negative contempt for reaching near zero position value if he can do it.
Obviously in the case of a figure disadvantage he can lessen the drawback if he can find a tactical opportunity.
I would be curious to know with a knight disadvantage the Stockfish is the better or the Komodo.
I think Stockfish is better near the same Elo as it is the difference between Stockfish and Komodo.
In the case of a Bishop disadvantage the situation is different because of the pair of Bishop.
I propose to Mr. Kaufman to run a match between Stockfish and Konodo with a knight odd.
I am also very curious to see the result even if I am not a believer of this type of matches.
-
- Posts: 5960
- Joined: Sun Jan 10, 2010 6:15 am
- Location: Maryland USA
Re: Certain top engines are only good at playing strong chess and that is all ?
If you mean having full strength SF and Komodo take turns giving each other knight odds, that would be a total waste of time. The side with the extra knight would win essentially every game, maybe one in a few hundred or so might somehow be a draw. I've run enough knight odds matches to know this. Now if the side with the extra knight had 1% of the time or so, that might be a worthwhile test. But the play of the engines at 1% of normal time might not be relevant either, unless it's something like 100' to 1' (plus 1% inc). But that takes too much computer time for me to run.corres wrote: ↑Sat May 30, 2020 8:49 pmI think it is only sense of a Stockfish-Komodo match, from at least 2 x 50 opening positions with alternated colors plus alternated odd, that is 4 x 50 = 200 games (at least).lkaufman wrote: ↑Sat May 30, 2020 8:03 pmEngines can only give knight odds to other engines when the rating difference is something like 800 elo (depending on time limit, threads, etc.). I did some tests a while ago in which I had both Komodo and Stockfish give knight odds to a much weaker engine (it might have been a crippled Komodo level, I forget the details) a while ago, one weak enough (like about Magnus Carlsen level!) to make for fairly close matches, and I recall the results of Stockfish and Komodo were very similar, but I don't remember enough details. I could redo the experiment at some point against some suitable engine, with high Contempt settings.corres wrote: ↑Sat May 30, 2020 7:07 pmStockfish strives to reduce the negative contempt for reaching near zero position value if he can do it.
Obviously in the case of a figure disadvantage he can lessen the drawback if he can find a tactical opportunity.
I would be curious to know with a knight disadvantage the Stockfish is the better or the Komodo.
I think Stockfish is better near the same Elo as it is the difference between Stockfish and Komodo.
In the case of a Bishop disadvantage the situation is different because of the pair of Bishop.
I propose to Mr. Kaufman to run a match between Stockfish and Konodo with a knight odd.
I am also very curious to see the result even if I am not a believer of this type of matches.
Komodo rules!