Difference between Computer and humans chess players

[Chessqueen]

So...
Seriously, I thought you were going to say one has fingers and one does not.

Well the biggest difference is that Computer engines do NOT CRY after losing a game

[lkaufman]

So...
Seriously, I thought you were going to say one has fingers and one does not.

Well the biggest difference is that Computer engines do NOT CRY after losing a game

I'm sure that ChessBase (for example) could make Fritz "Cry" after losing if enough customers requested that feature!

[Chessqueen]

So...
Seriously, I thought you were going to say one has fingers and one does not.

Well the biggest difference is that Computer engines do NOT CRY after losing a game

I'm sure that ChessBase (for example) could make Fritz "Cry" after losing if enough customers requested that feature!

Like I mentioned in the initial post that between Engines it is pretty much +50 or -50 and an engine rated 2714 more likely will not lose against a 2612 rated engines, but between humans GMs it could be 100 or sometimes 150 rating difference and nothing is guaranteed for the higher rated player, like in this case , Young talented 16 years old Pragg rated 2612 is outplaying his 2714 rated opponent
https://live.chessbase.com/en/watch/84t ... ters-2022/

[Chessqueen]

Most computer chess engines have a constant chess rating that can be within 50 Elo higher or lower, but if you take for instance a human GM rated around 2650 to 2700, their rating could be up or down several hundreds Elo, for instance in one game they might perform around 2200 or lower if they blunder and in other games they might perform around 2800 or even higher.

I disagree.

What is the definition of performance in a game?
There is one definition based on result and you can say that if you lose against 2600 you get a performance of 2200 and if you win against 2600 you get a performance of 3000 and in this case computers with rating 2600 may also win against 2600 in a single game and lose against 2600 in a single game.

Second possible definition is based on the quality of the moves and in this case the question is how you measure quality of the moves.
It is possible that in one opening some engine with 2600 has quality of moves that is equivalent to 2200 humans and in another opening the quality of the moves of it is better than carlsen.

2600 chess engines may avoid tactical blunders assuming no bugs that is not always the case but
they can do big positional blunders in some openings that allow relatively weak players to beat them if they prepare to play the right openings.

I can add that in human games it is not something common to see 2650 to 2700 players lose against 2300 or 2200.
Blunders are not something random and usually the stronger player play moves that increase the chances of the weaker player to blunder.

Here is a game where a 1966 rated player from Italy was crushing Niger Short rated 2633, until he became careless at the end and blundered. Now if it was a 1960 rated engine, it would have finished him off, without blundering

[Chessqueen]

Most computer chess engines have a constant chess rating that can be within 50 Elo higher or lower, but if you take for instance a human GM rated around 2650 to 2700, their rating could be up or down several hundreds Elo, for instance in one game they might perform around 2200 or lower if they blunder and in other games they might perform around 2800 or even higher.

I disagree.

What is the definition of performance in a game?
There is one definition based on result and you can say that if you lose against 2600 you get a performance of 2200 and if you win against 2600 you get a performance of 3000 and in this case computers with rating 2600 may also win against 2600 in a single game and lose against 2600 in a single game.

Second possible definition is based on the quality of the moves and in this case the question is how you measure quality of the moves.
It is possible that in one opening some engine with 2600 has quality of moves that is equivalent to 2200 humans and in another opening the quality of the moves of it is better than carlsen.

2600 chess engines may avoid tactical blunders assuming no bugs that is not always the case but
they can do big positional blunders in some openings that allow relatively weak players to beat them if they prepare to play the right openings.

I can add that in human games it is not something common to see 2650 to 2700 players lose against 2300 or 2200.
Blunders are not something random and usually the stronger player play moves that increase the chances of the weaker player to blunder.

Here is a game where a 1966 rated player from Italy was crushing Niger Short rated 2633, until he became careless at the end and blundered. Now if it was a 1960 rated engine, it would have finished him off, without blundering

I wanted to found out if a 2633 rated player can beat a 1966 player with this position taken from the game played between the italian 1966 rated player, so I got Snowy rated around 1966 vs Komodo Dragon at UCI_Elo = 2633, and of Corse Komodo Won easy, which tell me Niger Short played bad until reaching this position

[pgn][Event "Computer chess game"]
[Site "MININT-UB2PIMJ"]
[Date "2022.02.08"]
[Round "?"]
[White "Snowy_0_2_x64"]
[Black "Dragon-2.6.1-64bit-avx2"]
[Result "0-1"]
[BlackElo "2633"]
[Time "20:32:47"]
[WhiteElo "1966"]
[TimeControl "900+10"]
[SetUp "1"]
[FEN "2r1b3/p1q1k1p1/4p2p/2pnN3/8/2P3Q1/P4PPP/3RK2R b - - 0 1"]
[Termination "normal"]
[PlyCount "76"]
[WhiteType "program"]
[BlackType "program"]

1. ... Kf8 2. c4 Ne7 3. Kf1 Nf5 4. Qf4 Kg8 5. Qe4 Ba4 6. Rd2 Rb8 7. g3 Nd4
8. f4 Qa5 9. Qd3 Bc6 10. Nxc6 Nxc6 11. Rf2 Qa6 12. Kg2 Nd4 13. Rc1 Rb4 14.
Rc3 Ra4 15. a3 Qb7+ 16. Kh3 Ra6 17. g4 Rb6 18. g5 e5 19. fxe5 hxg5 20. Kg4
Rh6 21. Kg3 Qh1 22. Qd2 g4 23. Re3 Qc6 24. Kf4 Qg6 25. Qc3 Qf5+ 26. Kg3 Qh5
27. Rb2 Qh4+ 28. Kg2 Qxh2+ 29. Kf1 Qh1+ 30. Kf2 Rh5 31. Rg3 Rf5+ 32. Ke3
Qf1 33. Rbg2 Rxe5+ 34. Kd2 Qe1+ 35. Kd3 Qd1+ 36. Qd2 Qb1+ 37. Qc2 Nxc2 38.
Rxc2 Qb3+ 39. Kd2 {White resigns} *[/pgn]

[lkaufman]

Most computer chess engines have a constant chess rating that can be within 50 Elo higher or lower, but if you take for instance a human GM rated around 2650 to 2700, their rating could be up or down several hundreds Elo, for instance in one game they might perform around 2200 or lower if they blunder and in other games they might perform around 2800 or even higher.

I disagree.

What is the definition of performance in a game?
There is one definition based on result and you can say that if you lose against 2600 you get a performance of 2200 and if you win against 2600 you get a performance of 3000 and in this case computers with rating 2600 may also win against 2600 in a single game and lose against 2600 in a single game.

Second possible definition is based on the quality of the moves and in this case the question is how you measure quality of the moves.
It is possible that in one opening some engine with 2600 has quality of moves that is equivalent to 2200 humans and in another opening the quality of the moves of it is better than carlsen.

2600 chess engines may avoid tactical blunders assuming no bugs that is not always the case but
they can do big positional blunders in some openings that allow relatively weak players to beat them if they prepare to play the right openings.

I can add that in human games it is not something common to see 2650 to 2700 players lose against 2300 or 2200.
Blunders are not something random and usually the stronger player play moves that increase the chances of the weaker player to blunder.

Here is a game where a 1966 rated player from Italy was crushing Niger Short rated 2633, until he became careless at the end and blundered. Now if it was a 1960 rated engine, it would have finished him off, without blundering

I wanted to found out if a 2633 rated player can beat a 1966 player with this position taken from the game played between the italian 1966 rated player, so I got Snowy rated around 1966 vs Komodo Dragon at UCI_Elo = 2633, and of Corse Komodo Won easy, which tell me Niger Short played bad until reaching this position

[pgn][Event "Computer chess game"]
[Site "MININT-UB2PIMJ"]
[Date "2022.02.08"]
[Round "?"]
[White "Snowy_0_2_x64"]
[Black "Dragon-2.6.1-64bit-avx2"]
[Result "0-1"]
[BlackElo "2633"]
[Time "20:32:47"]
[WhiteElo "1966"]
[TimeControl "900+10"]
[SetUp "1"]
[FEN "2r1b3/p1q1k1p1/4p2p/2pnN3/8/2P3Q1/P4PPP/3RK2R b - - 0 1"]
[Termination "normal"]
[PlyCount "76"]
[WhiteType "program"]
[BlackType "program"]

1. ... Kf8 2. c4 Ne7 3. Kf1 Nf5 4. Qf4 Kg8 5. Qe4 Ba4 6. Rd2 Rb8 7. g3 Nd4
8. f4 Qa5 9. Qd3 Bc6 10. Nxc6 Nxc6 11. Rf2 Qa6 12. Kg2 Nd4 13. Rc1 Rb4 14.
Rc3 Ra4 15. a3 Qb7+ 16. Kh3 Ra6 17. g4 Rb6 18. g5 e5 19. fxe5 hxg5 20. Kg4
Rh6 21. Kg3 Qh1 22. Qd2 g4 23. Re3 Qc6 24. Kf4 Qg6 25. Qc3 Qf5+ 26. Kg3 Qh5
27. Rb2 Qh4+ 28. Kg2 Qxh2+ 29. Kf1 Qh1+ 30. Kf2 Rh5 31. Rg3 Rf5+ 32. Ke3
Qf1 33. Rbg2 Rxe5+ 34. Kd2 Qe1+ 35. Kd3 Qd1+ 36. Qd2 Qb1+ 37. Qc2 Nxc2 38.
Rxc2 Qb3+ 39. Kd2 {White resigns} *[/pgn]

I believe you got the story completely wrong. The 1966 player didn't "become careless and blunder", where did you get that from? All reports including the video you cite say that his cellphone went off and he was forfeited. Well, that is "careless", and I suppose you could call that a "blunder", but that's not how anyone would read this comment.

[Chessqueen]

Most computer chess engines have a constant chess rating that can be within 50 Elo higher or lower, but if you take for instance a human GM rated around 2650 to 2700, their rating could be up or down several hundreds Elo, for instance in one game they might perform around 2200 or lower if they blunder and in other games they might perform around 2800 or even higher.

I disagree.

What is the definition of performance in a game?
There is one definition based on result and you can say that if you lose against 2600 you get a performance of 2200 and if you win against 2600 you get a performance of 3000 and in this case computers with rating 2600 may also win against 2600 in a single game and lose against 2600 in a single game.

Second possible definition is based on the quality of the moves and in this case the question is how you measure quality of the moves.
It is possible that in one opening some engine with 2600 has quality of moves that is equivalent to 2200 humans and in another opening the quality of the moves of it is better than carlsen.

2600 chess engines may avoid tactical blunders assuming no bugs that is not always the case but
they can do big positional blunders in some openings that allow relatively weak players to beat them if they prepare to play the right openings.

I can add that in human games it is not something common to see 2650 to 2700 players lose against 2300 or 2200.
Blunders are not something random and usually the stronger player play moves that increase the chances of the weaker player to blunder.

Here is a game where a 1966 rated player from Italy was crushing Niger Short rated 2633, until he became careless at the end and blundered. Now if it was a 1960 rated engine, it would have finished him off, without blundering

I wanted to found out if a 2633 rated player can beat a 1966 player with this position taken from the game played between the italian 1966 rated player, so I got Snowy rated around 1966 vs Komodo Dragon at UCI_Elo = 2633, and of Corse Komodo Won easy, which tell me Niger Short played bad until reaching this position

[pgn][Event "Computer chess game"]
[Site "MININT-UB2PIMJ"]
[Date "2022.02.08"]
[Round "?"]
[White "Snowy_0_2_x64"]
[Black "Dragon-2.6.1-64bit-avx2"]
[Result "0-1"]
[BlackElo "2633"]
[Time "20:32:47"]
[WhiteElo "1966"]
[TimeControl "900+10"]
[SetUp "1"]
[FEN "2r1b3/p1q1k1p1/4p2p/2pnN3/8/2P3Q1/P4PPP/3RK2R b - - 0 1"]
[Termination "normal"]
[PlyCount "76"]
[WhiteType "program"]
[BlackType "program"]

1. ... Kf8 2. c4 Ne7 3. Kf1 Nf5 4. Qf4 Kg8 5. Qe4 Ba4 6. Rd2 Rb8 7. g3 Nd4
8. f4 Qa5 9. Qd3 Bc6 10. Nxc6 Nxc6 11. Rf2 Qa6 12. Kg2 Nd4 13. Rc1 Rb4 14.
Rc3 Ra4 15. a3 Qb7+ 16. Kh3 Ra6 17. g4 Rb6 18. g5 e5 19. fxe5 hxg5 20. Kg4
Rh6 21. Kg3 Qh1 22. Qd2 g4 23. Re3 Qc6 24. Kf4 Qg6 25. Qc3 Qf5+ 26. Kg3 Qh5
27. Rb2 Qh4+ 28. Kg2 Qxh2+ 29. Kf1 Qh1+ 30. Kf2 Rh5 31. Rg3 Rf5+ 32. Ke3
Qf1 33. Rbg2 Rxe5+ 34. Kd2 Qe1+ 35. Kd3 Qd1+ 36. Qd2 Qb1+ 37. Qc2 Nxc2 38.
Rxc2 Qb3+ 39. Kd2 {White resigns} *[/pgn]

I believe you got the story completely wrong. The 1966 player didn't "become careless and blunder", where did you get that from? All reports including the video you cite say that his cellphone went off and he was forfeited. Well, that is "careless", and I suppose you could call that a "blunder", but that's not how anyone would read this comment.

He was careless not to turn off his phone, but that is proof that among human players sometimes there could be a difference of up to 600 rating points and sometimes they perform much much higher whereas between engines the most that I have seen is probably an upset of 100 elo between one weaker engine and another stronger engine.

[Chessqueen]

I wanted to found out if a 2633 rated player can beat a 1966 player with this position taken from the game played between the italian 1966 rated player, so I got Snowy rated around 1966 vs Komodo Dragon at UCI_Elo = 2633, and of Course Komodo Won easy the first time, but not the 2nd time around, which tell me Nigel Short played bad until reaching this position. I decided to have the same game replayed between Snowy vs Komodo Dragon at UCI_Elo = 2633, I thought that Komodo Dragon 2.6.1 was going to replay the exact moves, but to my surprise, it chose different moves and it lost, can Mr. Kauman explain why replaying the exact position this happened. This was the original game played between an italian player rated 1966 and GM Nigel Short rated 2633

[pgn][Event "Computer chess game"]
[Site "MININT-UB2PIMJ"]
[Date "2022.02.09"]
[Round "2"]
[White "Snowy_0_2_x64"]
[Black "Dragon-2.6.1-64bit-avx2"]
[Result "1-0"]
[BlackElo "2633"]
[Time "07:54:14"]
[WhiteElo "1966"]
[TimeControl "900+10"]
[SetUp "1"]
[FEN "2r1b3/p1q1k1p1/4p2p/2pnN3/8/2P3Q1/P4PPP/3RK2R b K - 0 1"]
[Termination "normal"]
[PlyCount "115"]
[WhiteType "program"]
[BlackType "program"]

1. ... Kf8 2. O-O Kg8 3. Rfe1 Bh5 4. Rc1 Rf8 5. Nd3 c4 6. Qxc7 Nxc7 7. Ne5
Nd5 8. g4 Be8 9. Nxc4 h5 10. gxh5 Nf4 11. Ne5 Bxh5 12. Rc2 Rf5 13. Kf1 Ng6
14. Nc6 e5 15. Nxa7 Bg4 16. Re3 Rh5 17. Kg1 Nh4 18. Nc6 Bd1 19. Rc1 Ba4 20.
Nb4 Bd7 21. Rd1 Bg4 22. Rd5 Nf3+ 23. Kg2 Rh3 24. Rd8+ Kh7 25. Rf8 Nh4+ 26.
Kf1 Ng6 27. Rb8 Rxh2 28. Kg1 Rh5 29. f3 Rh3 30. Kf2 Be6 31. a4 Nh4 32. Nd3
Rh2+ 33. Kg1 Rg2+ 34. Kh1 Ra2 35. Rb4 Nf5 36. Rxe5 Bc8 37. Kg1 Ra1+ 38. Kf2
Ra2+ 39. Ke1 Ra1+ 40. Ke2 Ra2+ 41. Kd1 Nd6 42. Rh5+ Kg8 43. Rd5 Nb7 44. Nf4
Ra3 45. Rc4 Bf5 46. Kd2 Ra1 47. Rc7 Rxa4 48. Ke3 Ra8 49. Rxf5 Nd6 50. Rg5
Ne8 51. Re7 Rc8 52. c4 Ra8 53. Ne6 Ra1 54. Kf4 Nf6 55. Rgxg7+ Kh8 56. Rgf7
Ra4 57. Rxf6 Rxc4+ 58. Kg3 Rg4+ {Black resigns} *[/pgn]

[lkaufman]

I wanted to found out if a 2633 rated player can beat a 1966 player with this position taken from the game played between the italian 1966 rated player, so I got Snowy rated around 1966 vs Komodo Dragon at UCI_Elo = 2633, and of Course Komodo Won easy the first time, but not the 2nd time around, which tell me Nigel Short played bad until reaching this position. I decided to have the same game replayed between Snowy vs Komodo Dragon at UCI_Elo = 2633, I thought that Komodo Dragon 2.6.1 was going to replay the exact moves, but to my surprise, it chose different moves and it lost, can Mr. Kauman explain why replaying the exact position this happened. This was the original game played between an italian player rated 1966 and GM Nigel Short rated 2633

[pgn][Event "Computer chess game"]
[Site "MININT-UB2PIMJ"]
[Date "2022.02.09"]
[Round "2"]
[White "Snowy_0_2_x64"]
[Black "Dragon-2.6.1-64bit-avx2"]
[Result "1-0"]
[BlackElo "2633"]
[Time "07:54:14"]
[WhiteElo "1966"]
[TimeControl "900+10"]
[SetUp "1"]
[FEN "2r1b3/p1q1k1p1/4p2p/2pnN3/8/2P3Q1/P4PPP/3RK2R b K - 0 1"]
[Termination "normal"]
[PlyCount "115"]
[WhiteType "program"]
[BlackType "program"]

1. ... Kf8 2. O-O Kg8 3. Rfe1 Bh5 4. Rc1 Rf8 5. Nd3 c4 6. Qxc7 Nxc7 7. Ne5
Nd5 8. g4 Be8 9. Nxc4 h5 10. gxh5 Nf4 11. Ne5 Bxh5 12. Rc2 Rf5 13. Kf1 Ng6
14. Nc6 e5 15. Nxa7 Bg4 16. Re3 Rh5 17. Kg1 Nh4 18. Nc6 Bd1 19. Rc1 Ba4 20.
Nb4 Bd7 21. Rd1 Bg4 22. Rd5 Nf3+ 23. Kg2 Rh3 24. Rd8+ Kh7 25. Rf8 Nh4+ 26.
Kf1 Ng6 27. Rb8 Rxh2 28. Kg1 Rh5 29. f3 Rh3 30. Kf2 Be6 31. a4 Nh4 32. Nd3
Rh2+ 33. Kg1 Rg2+ 34. Kh1 Ra2 35. Rb4 Nf5 36. Rxe5 Bc8 37. Kg1 Ra1+ 38. Kf2
Ra2+ 39. Ke1 Ra1+ 40. Ke2 Ra2+ 41. Kd1 Nd6 42. Rh5+ Kg8 43. Rd5 Nb7 44. Nf4
Ra3 45. Rc4 Bf5 46. Kd2 Ra1 47. Rc7 Rxa4 48. Ke3 Ra8 49. Rxf5 Nd6 50. Rg5
Ne8 51. Re7 Rc8 52. c4 Ra8 53. Ne6 Ra1 54. Kf4 Nf6 55. Rgxg7+ Kh8 56. Rgf7
Ra4 57. Rxf6 Rxc4+ 58. Kg3 Rg4+ {Black resigns} *[/pgn]

The Elo levels on Dragon have significant variety settings, both to weaken the play and to insure variety from one game to next, so you should get different moves. Also note that Snowy is about 1966 CCRL, not 1966 FIDE, it is much stronger than 1966 FIDE (though apparently so was the human player in this game).