Lyudmil Tsvetkov wrote:Let me post here some flattering remarks about a game of Komodo, so that Don does not decide I am criticising engines, and especially Komodo all the way. I am appreciating Don's open-minded approach and, as he posts frequently in my threads, Komodo definitely deserves special treating.
There is no need to flatter me but I appreciate the thought!
I don't have any problems with you criticizing the play of Komodo - in fact I always pay attention to comments about Komodo as I'm eager to discover problems that I might be able to fix for an ELO improvement.
I was only pointing out that I take criticism of computer play with a grain of salt as it very often turns out that the computer is right and the human is wrong. But not always.
Do you realize that all the top players now use computers to check their analysis? I am not a very strong player but with the help of the computer I can analyze positions and find errors and such. I'm surprise you don't use them for checking variations - you really should. You don't have to agree with the computer but like a good adviser they are there to advise, not to dictate.
Capital punishment would be more effective as a preventive measure if it were administered prior to the crime.
[D]r2qr1k1/2p1bpp1/p4n1p/2pp1P2/5BP1/2N2Q2/PPP1RP1P/R5K1 b - - 0 16
This is a critical position of the 62nd game, Houdini-Komodo.
Both engines currently think the game is perfectly equal, but Komodo now plays 16...d4 and, in less than a dozen moves already sees itself half a pawn down, not to mention Houdini's score. Of course, the move is very off target, as black is left with the semblance of double horizontally isolated pawns (c5 and c7, as the only pawn on adjacent file, d4, is in front of the more advanced double pawn), completely unnecessary bearing in mind that black actually has much better choices.
Although both engines think the position before d4 is fully even, I think black probably has an observable advantage, do not know if decisive.
The point is that the white double pawns on the king side are part of symmetrical pawn structures, meaning that those double pawns are a major liability, much bigger than a normal double pawn. The penalty might be somewhere twice the penalty for a normal double pawn, part of a group, and commensurate with the penalty for double horizontally isolated pawns. A further eval trick might consider that the penalty for those pawns is not due because one of them has already crossed the center line, but a further deeper look would ascertain that it is actually due, as those pawns have big difficulty to advance, because black is able to efficiently control all squares in front of them. This would mean that the pawn that gains space advantage (f5) exhibits just a semblance of importance and is there to stay out of job. Of course, one of the reasons for the excellent black piece control in front of those pawns is the good black pawn center, preventing enemy pieces from intervening.
Thus, I think black is much better here, and the exchange of the white knight for the black bishop on f5 earlier on has been wrong.
Increasing the advantage, however, is a different story, as tactical mistakes should be avoided. A plan including Qd7, c6 (or Bd6 first, depending on the white response), Bd6, exchanging the heavy pieces, seems a natural one to me, but also c4 might be considered in some lines. In this way, while the white pawn majority on the king side would not be felt at all, the black pawn majority on the queen side would, and this already swings the balance.
Another important thing to note is that the penalties for all weak pawns (double, isolated, etc.) should as a rule be increased with the presence of minor pieces on the board (as they are easier to exploit then, similarly to simple pawn endgames), but decreased with the presence of heavies (as those tend to level out existing weaknesses). Thus, the exchange of the heavy pieces would only favour black that has better pawn structure. An useful rule of thumb might be that for each heavy piece on the board weak pawns are scored down by a percentage point, while for each minor piece they are scored up by the same percentage point.
I know many will doubt the validity of what is written here, and the assessment of the position, as there might be some tactical details still to consider, but I think black is considerably better on the main diagram. In any case, whatever games I have played with this line with black, both against humans and engines, black has fared significantly better.
Any comments or additional engine output very much appreciated.
Lyudmil Tsvetkov wrote:[D]1r2r1k1/1pn2p1p/2pb1qp1/p4p2/3P4/1QP3P1/PP3PBP/R1B2RK1 w - - 0 17
This is a position from the 57th game of the tournament.
Stockfish has played very well the opening with white against Houdini, but now decides to attack and eat with the queen an enemy pawn far from where the main battle takes place. Stockfish plays now 17.Qa4, upon which Houdini replies with 17...f4, takes the initiative and builds up a devastating king attack.
I think Stockfish could easily keep some edge by playing 17.f4 itself instead, depriving black of any attacking opportunities and using its majority pawn on the queen side, when the time is favourable for that. I think playing 17.f4 is a must in this position, even if black did not threaten with a very menacing attack, because this move would be very helpful in the long term: it fixes (blocks) the most advanced black double pawn (f5) with an own pawn, and that is a serious advantage, long-term at that, that should be worth some 10-15cps. In this way the weakness of the double pawn becomes difficult or almost impossible to straighten, and that is always a positional asset. I think that whenever a real possibility to block with an own pawn the most advanced enemy double pawn exists, such a move should enjoy priority status, because a move or 2 later it could already not be playable, thus missing a valuable positional edge.
Of course, white might have been afraid that f4 would leave the e3 square undefended, but that is much less important in this particular case.
Any engines there willing to play 17.f4?
Best, Lyudmil
My little Amateur (slightly updated) will play f4. Not sure it's the best move, but at least it's human-like.
Lyudmil Tsvetkov wrote:Let me post here some flattering remarks about a game of Komodo, so that Don does not decide I am criticising engines, and especially Komodo all the way. I am appreciating Don's open-minded approach and, as he posts frequently in my threads, Komodo definitely deserves special treating.
There is no need to flatter me but I appreciate the thought!
I don't have any problems with you criticizing the play of Komodo - in fact I always pay attention to comments about Komodo as I'm eager to discover problems that I might be able to fix for an ELO improvement.
I was only pointing out that I take criticism of computer play with a grain of salt as it very often turns out that the computer is right and the human is wrong. But not always.
Do you realize that all the top players now use computers to check their analysis? I am not a very strong player but with the help of the computer I can analyze positions and find errors and such. I'm surprise you don't use them for checking variations - you really should. You don't have to agree with the computer but like a good adviser they are there to advise, not to dictate.
Hi Don.
The above-mentioned Komodo game was really great.
Computers are excellent in helping you see a large variety of lines you would not suppose they would exist, and, very often, when I start losing a game, and the result is already clear, I will go back to the game with computer suggestions to see something different and intriguing. Many lines I would completely miss.
But, computers also misplay a wide variety of position types, and any human might take advantage of that. As with humans, human-engine struggle is about who is going to impose her playing style on the other party. Whoever does that will usually be favoured in some way. Therefore, I choose against computers strictly only lines they do not understand. But when I want to learn something, I start looking into engine lines.
After looking a bit more carefully into the game, I will agree that my suggested continuation Ng4 is not sufficient, but probably also the whole sacrifice would prove insufficient. Black could have retained a small edge after the natural a5 instead of sacrificing. But it provided some food for thought.
I am grateful to iCE for its output that opened my eyes. It is always good to have some useful feedback. I would always appreciate a bit sharper comments, whenever those are substantiated by output/analysis.
Lyudmil Tsvetkov wrote:Let me post here some flattering remarks about a game of Komodo, so that Don does not decide I am criticising engines, and especially Komodo all the way. I am appreciating Don's open-minded approach and, as he posts frequently in my threads, Komodo definitely deserves special treating.
There is no need to flatter me but I appreciate the thought!
I don't have any problems with you criticizing the play of Komodo - in fact I always pay attention to comments about Komodo as I'm eager to discover problems that I might be able to fix for an ELO improvement.
I was only pointing out that I take criticism of computer play with a grain of salt as it very often turns out that the computer is right and the human is wrong. But not always.
Do you realize that all the top players now use computers to check their analysis? I am not a very strong player but with the help of the computer I can analyze positions and find errors and such. I'm surprise you don't use them for checking variations - you really should. You don't have to agree with the computer but like a good adviser they are there to advise, not to dictate.
Hi Don.
The above-mentioned Komodo game was really great.
Computers are excellent in helping you see a large variety of lines you would not suppose they would exist, and, very often, when I start losing a game, and the result is already clear, I will go back to the game with computer suggestions to see something different and intriguing. Many lines I would completely miss.
But, computers also misplay a wide variety of position types, and any human might take advantage of that. As with humans, human-engine struggle is about who is going to impose her playing style on the other party. Whoever does that will usually be favoured in some way. Therefore, I choose against computers strictly only lines they do not understand. But when I want to learn something, I start looking into engine lines.
But that's not how you have to use computers. When you see something that you think is the way to proceed just use the computer to see if there is a relatively simple refutation and it will save you a lot of time. Don't use it to dictate what move you should play. If it suggests a move that is interesting see if you can refute it! That is the best way to utilize computers.
Computers will find tactical errors in your analysis - but don't use them for telling you what you should play unless you can see for yourself that it's good.
Or do you simply fear that it will make you lazy? I think all the GM's are using computers in the way I am saying.
Capital punishment would be more effective as a preventive measure if it were administered prior to the crime.
Lyudmil Tsvetkov wrote:Let me post here some flattering remarks about a game of Komodo, so that Don does not decide I am criticising engines, and especially Komodo all the way. I am appreciating Don's open-minded approach and, as he posts frequently in my threads, Komodo definitely deserves special treating.
There is no need to flatter me but I appreciate the thought!
I don't have any problems with you criticizing the play of Komodo - in fact I always pay attention to comments about Komodo as I'm eager to discover problems that I might be able to fix for an ELO improvement.
I was only pointing out that I take criticism of computer play with a grain of salt as it very often turns out that the computer is right and the human is wrong. But not always.
Do you realize that all the top players now use computers to check their analysis? I am not a very strong player but with the help of the computer I can analyze positions and find errors and such. I'm surprise you don't use them for checking variations - you really should. You don't have to agree with the computer but like a good adviser they are there to advise, not to dictate.
Hi Don.
The above-mentioned Komodo game was really great.
Computers are excellent in helping you see a large variety of lines you would not suppose they would exist, and, very often, when I start losing a game, and the result is already clear, I will go back to the game with computer suggestions to see something different and intriguing. Many lines I would completely miss.
But, computers also misplay a wide variety of position types, and any human might take advantage of that. As with humans, human-engine struggle is about who is going to impose her playing style on the other party. Whoever does that will usually be favoured in some way. Therefore, I choose against computers strictly only lines they do not understand. But when I want to learn something, I start looking into engine lines.
But that's not how you have to use computers. When you see something that you think is the way to proceed just use the computer to see if there is a relatively simple refutation and it will save you a lot of time. Don't use it to dictate what move you should play. If it suggests a move that is interesting see if you can refute it! That is the best way to utilize computers.
Computers will find tactical errors in your analysis - but don't use them for telling you what you should play unless you can see for yourself that it's good.
Or do you simply fear that it will make you lazy? I think all the GM's are using computers in the way I am saying.
Hi Don.
Thanks God I am not a grandmaster and I do not compete. My first care would be to avoid dull routine.
What is the use of computer utilisation, if all those grandmasters are unable to defeat top engines of today on big hardware? You say, humans should leave the supremacy to computers, but I do not agree. I think there is no reason at all that the human brain should be in any way deficient to engine strength, even if supported by vast hardware.
People just would not care to find their own way of play that would be menacing to computers, they would follow computer recommendations and the dull routine instead. I do not imagine a world where you are resigned that something is already indisputably better than you, and you give up all fight. This is not real life, but a surrogate of life.
Thanks God I am not a grandmaster and I do not compete. My first care would be to avoid dull routine.
What is the use of computer utilisation, if all those grandmasters are unable to defeat top engines of today on big hardware? You say, humans should leave the supremacy to computers, but I do not agree.
I didn't say that! Didn't you read what I just posted?
I basically said that you should do your own thinking and that computer could be relegated to the role of servant. It could simply do the grunt work while YOU do the creative thinking.
That is exactly how the top players use computers. Use them to double check your variations for obvious tactical blunders - etc ...
I think there is no reason at all that the human brain should be in any way deficient to engine strength, even if supported by vast hardware.
People just would not care to find their own way of play that would be menacing to computers, they would follow computer recommendations and the dull routine instead. I do not imagine a world where you are resigned that something is already indisputably better than you, and you give up all fight. This is not real life, but a surrogate of life.
Capital punishment would be more effective as a preventive measure if it were administered prior to the crime.
The 6th game of the 3Champs was a real joy for me. I would award it all possible prizes, even if there might have been some mistakes on both parts. The game was a prolonged battle, and you could find in it unusual opening, interesting pawn structure, opposite-colour bishops, advance of passers, sharp attacks and counter-attacks, modernistic understanding of shelters with far advanced pawns, closed pawns chains, reminiscences of space advantage, good prophylactic moves, many vivid, picturesque tableaus, some shuffling, and, of course, a lot suspense!
I think this is how computer chess needs to proceed in the future, and it should have an awful lot of supporters.
