An idea for new Handicap games for dragon

[Chessqueen]

hikaru did not win a single game in the last match.
The interesting question about error rate is what is the error rate in case the handicap is big enough for the human to win every game.

I do not know which tool to use to calculate the error rate but every human who know can test it by himself by choosing handicap that is big enough for him to win every game.

He should have won 1 game, as noted by the analysis. Look at missed wins.

"Missed win: Black=1."

Most of the games show Black missing a -+ decisive advantage at some point, presumably "missed win" just has an even higher threshold score. But this suggests that if the score is beyond whatever is considered a "win", bad moves which leave the score above the win may not be called mistakes. Probably knight odds is near or above the winning score line, so this might mean that the analysis just isn't useful when the handicap is a knight or more. Is there any easy way to determine if this is the case or not?

I was referring to Centipawn analysis scoring. 2 of the 8 games was showing Black missed a win. And how many time it occurred in the game. Of the 2 games it happened once in each of those games.

I don't know how GM Carlsen, Andrew Tang and GM Nakamura can play Bullet Chess at lightning speed and only make a few blunders per game. I Never ever play Bullet but I tried today versus the Free Version of Komodo version 12.1 with a Rook and got totally destroyed, I blundered too often at that speed, then I tried at Rook plus b2 pawns Odds which is way too much if I play in game 20 Minutes plus 10 sec increment, but at Bullet I could hardly move my pieces fast enough and every time I either blunder a pawn, a Knight or a Bishop, it is so hard to see all the tactical threat in a coupe of seconds. Anyway these people are super Human, but still too weak to play versus any engines above 3400 in Bullet, since engine do NOT blunder. Speaking about GM Nakamura did he HIT the lottery? His Net worth is $50 millions much much more the the champions GM Carlsen
https://networthplanet.com/athletes/hik ... he%20world.

https://www.wealthypersons.com/hikaru-n ... 2020-2021/

Compared to Carlsen Net worth of only 8 million and being the chess Champion https://www.celebritynetworth.com/riche ... net-worth/

Nakamura made a reference to this claim during our match (he quoted $45 million) in a way that implied that it was total nonsense. Not everything on the internet is true.

You are correct do NOT believe everything that is on the internet

[mwyoung]

hikaru did not win a single game in the last match.
The interesting question about error rate is what is the error rate in case the handicap is big enough for the human to win every game.

I do not know which tool to use to calculate the error rate but every human who know can test it by himself by choosing handicap that is big enough for him to win every game.

I do not know as I have tested every one I could find except this game. This is the exact position ChessQueen won. And a GM was only able to draw. And the GM also had a human error rate unlike ChessQueen.

Code: Select all

KomodoChess:   15  => Average=0.15
gmjoel:   29  => Average=0.29

[pgn][Event "Live Chess - Odds Chess"]
[Site "Chess.com"]
[Date "2016.03.23"]
[Round "?"]
[White "KomodoChess"]
[Black "gmjoel"]
[Result "1/2-1/2"]
[WhiteElo "3368"]
[BlackElo "2200"]
[Annotator "Komodo 9.01 64-bit"]
[SetUp "1"]
[FEN "rn1qk1nr/pppppppp/8/8/8/8/PPPPPPPP/RNB1KBNR w KQkq - 0 1"]
[PlyCount "88"]
[EventDate "2016.??.??"]
[TimeControl "2700+30"]

1. d4 {Dragon by Komodo Chess 64-bit:} Nf6 2. e3 Nc6 3. Nf3 {OK} O-O 4. a3 d6
5. Be2 {OK} e5 6. O-O {OK} exd4 7. exd4 Re8 8. Nc3 Qd7 9. Be3 Ne7 {OK} 10. Bd3
{OK} Ned5 11. Nxd5 {OK} Nxd5 12. Bd2 {OK} h6 13. c4 {OK} Nf6 {OK} 14. d5 Ne4
15. Be3 {OK} a5 16. Nd4 {OK} a4 {OK} 17. Rab1 Re7 18. Bc2 Rae8 19. h3 Nc5 20.
Bd2 {OK} Ne4 21. Bf4 Nc5 22. Bd2 Re2 {This clearly cannot be the right move,
but Benjamin could find no way to make progress. Benjamin cleverly nabs a pawn
in the resulting endgame, but progress from there is hard to come by.} 23. Bf5
{OK} Qxf5 24. Nxf5 Rxd2 {OK} 25. Ne3 {OK} Re4 26. Rfd1 Red4 {Mistake (1.0)} (
26... Rxd1+ $19) 27. Rbc1 {\"It drew with this idea of Rc1 and giving up the
b2-pawn. It was not obvious at all.\" - Benjamin. Mistake (1.0)} (27. Kf1 $17)
27... g6 28. Rxd2 {OK} Rxd2 29. Rc2 {Best} Rxc2 {OK} 30. Nxc2 Nd3 {OK} 31. Kf1
{Mistake (1.1)} (31. g3 $142) 31... Nxb2 32. Ne3 {Best} f5 33. g3 {Best} Nd3 {
OK} 34. Ke2 Nc5 35. Nd1 {OK} Kf7 {OK} 36. Ke3 {Mistake (0.8)} (36. Nc3 $17)
36... c6 37. dxc6 bxc6 38. Nc3 Ke6 {Mistake (1.2)} (38... g5 $19) 39. Ne2 $1
$15 {Strong} Kd7 {Inaccurate} 40. Nf4 {OK} g5 {OK} 41. Ne2 {[#] OK} Nb3 {
Mistake (1.2)} (41... h5 $1 $17) 42. Nc3 Nc5 {OK} 43. Ne2 {[#] Mistake (0.9)} (
43. f4 $15) 43... Nb3 {Mistake (1.2)} (43... h5 $1 $17) 44. Nc3 $11 {OK} Nc5 {
Game drawn by agreement. Strong: White=1 Best: White=3 OK: White=17, Black=10
Inaccurate: Black=1 Mistake: White=4, Black=4. Weighted Error Value: White=0.
15/Black=0.29 Centipawn loss: w=15/b=25} 1/2-1/2

[/pgn]

[Chessqueen]

hikaru did not win a single game in the last match.
The interesting question about error rate is what is the error rate in case the handicap is big enough for the human to win every game.

I do not know which tool to use to calculate the error rate but every human who know can test it by himself by choosing handicap that is big enough for him to win every game.

Use this position and figure it out yuourself
Uri is correct the more pieced you take the less you can blunder, I wonder what is the average error rate in the following position setup, so can you beat Komodo 9 with Black, mwyoung

[d]rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQK3 w Qkq - 0 1']

[mwyoung]

hikaru did not win a single game in the last match.
The interesting question about error rate is what is the error rate in case the handicap is big enough for the human to win every game.

I do not know which tool to use to calculate the error rate but every human who know can test it by himself by choosing handicap that is big enough for him to win every game.

Use this position and figure it out yuourself
Uri is correct the more pieced you take the less you can blunder, I wonder what is the average error rate in the following position setup, so can you beat Komodo 9 with Black, mwyoung

[d]rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQK3 w Qkq - 0 1']

Fiction can be fun. But I find the reference section more enlightening. And this is your exact position you WON! But a GM was only able to DRAW with both you and the GM playing Komodo. And unlike you! The GM had a human error rate. And you were shown to be using a computer to win the game.

[Chessqueen]

hikaru did not win a single game in the last match.
The interesting question about error rate is what is the error rate in case the handicap is big enough for the human to win every game.

I do not know which tool to use to calculate the error rate but every human who know can test it by himself by choosing handicap that is big enough for him to win every game.

Use this position and figure it out yourself
Uri is correct the more pieced you take the less you can blunder, I wonder what is the average error rate in the following position setup, so can you beat Komodo 9 with Black, mwyoung

Well since I gave you an easy one to play and you did NOT want to show us the centipawns theory, I will demonstrate it to you how it is done, I was expecting 10.Be3 but I took Komodo resignation. Please tell me what is the centipawn score here. To me when Komodo has given lot of material it play really stupid and your theory is useless. Please tell me why Komodo played 8.Na2? in this position are you going to blame it on my refurbished computer or because I do NOT have at least 16 MB of memory, come on I am waiting for your theory? https://www.officedepot.com/a/products/ ... lsrc=aw.ds

[pgn][Event "Blitz:5'+1""]
[Site "MyTown"]
[Date "2020.11.12"]
[Round "9"]
[White "Komodo 13.3 64-bit"]
[Black "ChessQueen"]
[Result "0-1"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQK3 w Qkq - 0 1"]
[PlyCount "48"]
[TimeControl "300+1"]

1. d4 b6 2. Nc3 Nf6 3. Bf4 Bb7 4. Qd2 Bxg2 5. O-O-O e6 6. Kb1 Nc6 7. a3 d5 8.
Na2 Ne4 9. Qe1 Qf6
0-1[/pgn]

[Chessqueen]

hikaru did not win a single game in the last match.
The interesting question about error rate is what is the error rate in case the handicap is big enough for the human to win every game.

I do not know which tool to use to calculate the error rate but every human who know can test it by himself by choosing handicap that is big enough for him to win every game.

Use this position and figure it out yourself
Uri is correct the more pieced you take the less you can blunder, I wonder what is the average error rate in the following position setup, so can you beat Komodo 9 with Black, mwyoung

Well since I gave you an easy one to play and you did NOT want to show us the centipawns theory, I will demonstrate it to you how it is done, I was expecting 10.Be3 but I took Komodo resignation. Please tell me what is the centipawn score here. To me when Komodo has given lot of material it play really stupid and your theory is useless. Please tell me why Komodo played 8.Na2? in this position are you going to blame it on my refurbished computer or because I do NOT have at least 16 MB of memory, come on I am waiting for your theory? That position is so lost for Komodo, that I was even considering to take 8.....Nxd4 followed by 9 Bc5 ! exchanging two pawns for the Knight even if the Queen move to e5! I was really considering doing it https://www.officedepot.com/a/products/ ... lsrc=aw.ds

[pgn][Event "Blitz:5'+1""]
[Site "MyTown"]
[Date "2020.11.12"]
[Round "9"]
[White "Komodo 13.3 64-bit"]
[Black "ChessQueen"]
[Result "0-1"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQK3 w Qkq - 0 1"]
[PlyCount "48"]
[TimeControl "300+1"]

1. d4 b6 2. Nc3 Nf6 3. Bf4 Bb7 4. Qd2 Bxg2 5. O-O-O e6 6. Kb1 Nc6 7. a3 d5 8.
Na2 Ne4 9. Qe1 Qf6
0-1[/pgn]

[Chessqueen]

hikaru did not win a single game in the last match.
The interesting question about error rate is what is the error rate in case the handicap is big enough for the human to win every game.

I do not know which tool to use to calculate the error rate but every human who know can test it by himself by choosing handicap that is big enough for him to win every game.

Use this position and figure it out yourself
Uri is correct the more pieced you take the less you can blunder, I wonder what is the average error rate in the following position setup, so can you beat Komodo 9 with Black, mwyoung

Well since I gave you an easy one to play and you did NOT want to show us the centipawns theory, I will demonstrate it to you how it is done, I was expecting 10.Be3 but I took Komodo resignation. Please tell me what is the centipawn score here. To me when Komodo has given lot of material it play really stupid and your theory is useless. Please tell me why Komodo played 8.Na2? in this position are you going to blame it on my refurbished computer or because I do NOT have at least 16 MB of memory, come on I am waiting for your theory? That position is so lost for Komodo, that I was even considering to take 8.....Nxd4 followed by 9 Bc5 ! exchanging two pawns for the Knight even if the Queen move to e5! I was really considering doing it https://www.officedepot.com/a/products/ ... lsrc=aw.ds

[pgn][Event "Blitz:5'+1""]
[Site "MyTown"]
[Date "2020.11.12"]
[Round "9"]
[White "Komodo 13.3 64-bit"]
[Black "ChessQueen"]
[Result "0-1"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQK3 w Qkq - 0 1"]
[PlyCount "48"]
[TimeControl "300+1"]

1. d4 b6 2. Nc3 Nf6 3. Bf4 Bb7 4. Qd2 Bxg2 5. O-O-O e6 6. Kb1 Nc6 7. a3 d5 8.
Na2 Ne4 9. Qe1 Qf6
0-1[/pgn]

Here is another which Komodo does NOT have a clue of what to do, if I am alive 5 years from NOW I can still beat whatever version of Komodo with a Rook and a pawn Odds, guess what, I will be at least 200 rating points higher and I will still continue training with at least 500 games versus Komodo under my belt

[pgn][Event "Blitz:5'+1""]
[Site "MyTown"]
[Date "2020.11.12"]
[Round "10"]
[White "Komodo 13.3 64-bit"]
[Black "ChessQueen"]
[Result "0-1"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppp1p/8/8/8/8/PPPPPPPP/3QKBNR w Kkq - 0 1"]
[PlyCount "63"]
[TimeControl "300+1"]

1. g3 Bg7 2. h3 Bxb2 3. c3 Ba3 4. Qb3 Bd6 5. c4 e5 6. Qa4 Qf6 7. Bg2 Bc5 8. e3
Ne7 9. Nf3 O-O 10. O-O d5 11. cxd5 Nxd5 12. Qc2 Bd6 13. Rd1 Nb4 14. Qb1 Bf5 15.
d3 N8c6 16. Nd2 Bxd3 17. Ne4 Bxe4 18. Bxe4 Rab8 19. Bxh7+ Kg7 20. Be4 Qe6 21.
a3 Na6 22. Bf5 Qh6 23. Qe4 Bxa3 24. Ra1 Be7 25. Kh1 Nc5 26. Qg2 Bd6 27. Rf1 Ne7
28. Bb1 Qe6 29. Kh2 e4 30. Rh1 Qf5 31. Kg1 Qf3
0-1[/pgn]

[Uri Blass]

hikaru did not win a single game in the last match.
The interesting question about error rate is what is the error rate in case the handicap is big enough for the human to win every game.

I do not know which tool to use to calculate the error rate but every human who know can test it by himself by choosing handicap that is big enough for him to win every game.

I do not know as I have tested every one I could find except this game. This is the exact position ChessQueen won. And a GM was only able to draw. And the GM also had a human error rate unlike ChessQueen.

Code: Select all

KomodoChess:   15  => Average=0.15
gmjoel:   29  => Average=0.29

[pgn][Event "Live Chess - Odds Chess"]
[Site "Chess.com"]
[Date "2016.03.23"]
[Round "?"]
[White "KomodoChess"]
[Black "gmjoel"]
[Result "1/2-1/2"]
[WhiteElo "3368"]
[BlackElo "2200"]
[Annotator "Komodo 9.01 64-bit"]
[SetUp "1"]
[FEN "rn1qk1nr/pppppppp/8/8/8/8/PPPPPPPP/RNB1KBNR w KQkq - 0 1"]
[PlyCount "88"]
[EventDate "2016.??.??"]
[TimeControl "2700+30"]

1. d4 {Dragon by Komodo Chess 64-bit:} Nf6 2. e3 Nc6 3. Nf3 {OK} O-O 4. a3 d6
5. Be2 {OK} e5 6. O-O {OK} exd4 7. exd4 Re8 8. Nc3 Qd7 9. Be3 Ne7 {OK} 10. Bd3
{OK} Ned5 11. Nxd5 {OK} Nxd5 12. Bd2 {OK} h6 13. c4 {OK} Nf6 {OK} 14. d5 Ne4
15. Be3 {OK} a5 16. Nd4 {OK} a4 {OK} 17. Rab1 Re7 18. Bc2 Rae8 19. h3 Nc5 20.
Bd2 {OK} Ne4 21. Bf4 Nc5 22. Bd2 Re2 {This clearly cannot be the right move,
but Benjamin could find no way to make progress. Benjamin cleverly nabs a pawn
in the resulting endgame, but progress from there is hard to come by.} 23. Bf5
{OK} Qxf5 24. Nxf5 Rxd2 {OK} 25. Ne3 {OK} Re4 26. Rfd1 Red4 {Mistake (1.0)} (
26... Rxd1+ $19) 27. Rbc1 {\"It drew with this idea of Rc1 and giving up the
b2-pawn. It was not obvious at all.\" - Benjamin. Mistake (1.0)} (27. Kf1 $17)
27... g6 28. Rxd2 {OK} Rxd2 29. Rc2 {Best} Rxc2 {OK} 30. Nxc2 Nd3 {OK} 31. Kf1
{Mistake (1.1)} (31. g3 $142) 31... Nxb2 32. Ne3 {Best} f5 33. g3 {Best} Nd3 {
OK} 34. Ke2 Nc5 35. Nd1 {OK} Kf7 {OK} 36. Ke3 {Mistake (0.8)} (36. Nc3 $17)
36... c6 37. dxc6 bxc6 38. Nc3 Ke6 {Mistake (1.2)} (38... g5 $19) 39. Ne2 $1
$15 {Strong} Kd7 {Inaccurate} 40. Nf4 {OK} g5 {OK} 41. Ne2 {[#] OK} Nb3 {
Mistake (1.2)} (41... h5 $1 $17) 42. Nc3 Nc5 {OK} 43. Ne2 {[#] Mistake (0.9)} (
43. f4 $15) 43... Nb3 {Mistake (1.2)} (43... h5 $1 $17) 44. Nc3 $11 {OK} Nc5 {
Game drawn by agreement. Strong: White=1 Best: White=3 OK: White=17, Black=10
Inaccurate: Black=1 Mistake: White=4, Black=4. Weighted Error Value: White=0.
15/Black=0.29 Centipawn loss: w=15/b=25} 1/2-1/2

[/pgn]

It is the same position but the opponent was not the same.
The GM played a stronger opponent as larry kaufman showed.

When you play weaker opponent it is easier not to lose your winning advantage and I expect your error rate to be lower.

I dislike claiming that somebody cheated if I am not sure about it.

[Chessqueen]

Uri is correct the more pieced you take the less you can blunder, he will never rate our latest games and will continue bringing the same old position back again as his argument, this guy is like Trump he will never admit to lose an argument. I will admit that each one of those 5 games that I played I replayed them some 6 times or 7 times until I figure out how to beat Komodo, and made several notes of how I blundered and corrected my blunders in the final games then I saved those games and finally post3ed those 5 games, but for these last games which clearly I have a great advantage I played them only once and beat Komodo fair and square. So for the record I did what some people would probable do to show their best game against Komodo if they would have agreed to pay $100 or even $50.00 per game. Now that you know why and how I played so precise the true is revealed, but I had a purpose and it was to show ways how people could have won by playing over and over again the same positions, but not this simple and easy position which I am clearly winning nor the Rook Odds when I gave it one of my pawn .

[pgn][Event "Blitz:5'+1""]
[Site "MyTown"]
[Date "2020.11.12"]
[Round "10"]
[White "Komodo 13.3 64-bit"]
[Black "ChessQueen"]
[Result "0-1"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppp1p/8/8/8/8/PPPPPPPP/3QKBNR w Kkq - 0 1"]
[PlyCount "63"]
[TimeControl "300+1"]

1. g3 Bg7 2. h3 Bxb2 3. c3 Ba3 4. Qb3 Bd6 5. c4 e5 6. Qa4 Qf6 7. Bg2 Bc5 8. e3
Ne7 9. Nf3 O-O 10. O-O d5 11. cxd5 Nxd5 12. Qc2 Bd6 13. Rd1 Nb4 14. Qb1 Bf5 15.
d3 N8c6 16. Nd2 Bxd3 17. Ne4 Bxe4 18. Bxe4 Rab8 19. Bxh7+ Kg7 20. Be4 Qe6 21.
a3 Na6 22. Bf5 Qh6 23. Qe4 Bxa3 24. Ra1 Be7 25. Kh1 Nc5 26. Qg2 Bd6 27. Rf1 Ne7
28. Bb1 Qe6 29. Kh2 e4 30. Rh1 Qf5 31. Kg1 Qf3 White resigns} 0-1[/pgn]

[Chessqueen]

Uri is correct the more pieced you take the less you can blunder, he will never rate our latest games and will continue bringing the same old position back again as his argument, this guy is like Trump he will never admit to lose an argument. I will admit that each one of those 5 games that I played I replayed them some 6 times or 7 times until I figured out how to beat Komodo, and made several notes of how I blundered and corrected my blunders in the final games then I saved those games and finally posted those 5 games, but for these last games which clearly I have a great advantage I played them only once and beat Komodo fair and square. So for the record I did what some people would probably do to show their best game against Komodo if they would have agreed to pay $100 or even $50.00 per game. Now that you know why and how I played so precise the true is revealed, but I had a purpose and it was to show ways how people could have won by playing over and over again the same positions, but not this simple and easy position which I am clearly winning nor the Rook Odds when I gave it one of my pawn, Now that my purpose was shown and known to everybody here, lets continue the handicap experiment but with one correction at the moment that a person agrees to play a setup position he or she must accept it from probably 8 positions given at the moment and the game should be played on chess.com. If GM Nakamura is given the exact last four positions that he lost the 2nd day, he will at least draw them all

[pgn][Event "Blitz:5'+1""]
[Site "MyTown"]
[Date "2020.11.12"]
[Round "10"]
[White "Komodo 13.3 64-bit"]
[Black "ChessQueen"]
[Result "0-1"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppp1p/8/8/8/8/PPPPPPPP/3QKBNR w Kkq - 0 1"]
[PlyCount "63"]
[TimeControl "300+1"]

1. g3 Bg7 2. h3 Bxb2 3. c3 Ba3 4. Qb3 Bd6 5. c4 e5 6. Qa4 Qf6 7. Bg2 Bc5 8. e3
Ne7 9. Nf3 O-O 10. O-O d5 11. cxd5 Nxd5 12. Qc2 Bd6 13. Rd1 Nb4 14. Qb1 Bf5 15.
d3 N8c6 16. Nd2 Bxd3 17. Ne4 Bxe4 18. Bxe4 Rab8 19. Bxh7+ Kg7 20. Be4 Qe6 21.
a3 Na6 22. Bf5 Qh6 23. Qe4 Bxa3 24. Ra1 Be7 25. Kh1 Nc5 26. Qg2 Bd6 27. Rf1 Ne7
28. Bb1 Qe6 29. Kh2 e4 30. Rh1 Qf5 31. Kg1 Qf3 White resigns} 0-1[/pgn]