How many of you believe that Stockfish 12 or Komodo can give a rook Odds to a 1450 engine ?

[Chessqueen]

a bit more info about chess handicaps
https://en.wikipedia.org/wiki/Handicap_(chess)
here's a bit more from the historical perspective:
http://www.edochess.ca/batgirl/odds-giving.html

Uri Blass claimed in this forum he can beat most
chess players with queen plus double rook odds.
(but then includes unrated players). well there are|
players who hardly know the rules, so it might be true.

As for ratings vs material odds, it's a better idea to look
at rating ratio (vs difference) so Nakamura vs 1400 is
2.0 ratio, but that's also for a 1200 vs 600 player. Queens odds?
maybe , but that is approx 8.5 points and to get a more
balanced game maybe other material odds also can be
applied (eg. rook plus bishop odds instead of queen).
A matter of trial and error, but then again, accuracy isn't
so important i guess, because i think of a self-correcting
system (of rating ratio's). More simple would be just
differences for rating 'classes' as superGM, GM, IM, NM
A,B,C etc. players

I just found out that Stockfish latest version can NOT beat TTE rated around 1480 with a Queen Odds at TC 60'+1" Probably the latest version of Komodo can, since Larry Kaufman gave special instructions to play with Odds, but it has to be tested

[pgn][Event "Computer chess game"]
[Site "DESKTOP-OFQ3C0P"]
[Date "2021.02.11"]
[Round "?"]
[White "Stockfish_20090216_x64_modern"]
[Black "TarraschToyEngineV0.906"]
[Result "*"]
[BlackElo "1480"]
[Time "17:16:07"]
[WhiteElo "3480"]
[TimeControl "3600+5"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNB1KBNR w KQkq - 0 1"]
[Termination "normal"]
[PlyCount "9"]
[WhiteType "program"]
[BlackType "program"]

1. d4 {(d2-d4 e7-e6 Nb1-c3 Bf8-e7 Ng1-f3 d7-d6 e2-e4 Ng8-f6 Bc1-g5 Nb8-d7
e4-e5 d6xe5 Nf3xe5 Nd7xe5 d4xe5 Nf6-d7 Bg5xe7 Qd8xe7 O-O-O Nd7xe5 Rd1-e1
Ne5-g4 Re1-e2 O-O a2-a3 h7-h6 Nc3-e4) -12.24/31 110} d5 {(d7-d5 Nb1-c3
Nb8-c6 e2-e3 Ng8-f6 f2-f4 Bc8-f5 Bf1-d3) +8.90/8 60} 2. e3 {(e2-e3 e7-e6
Ng1-f3 c7-c6 c2-c4 Bf8-d6 Bf1-d3 Ng8-e7 Nb1-c3 Nb8-d7 O-O O-O Bc1-d2 a7-a6
Ra1-e1 Bd6-c7 Nc3-e2 h7-h6 c4xd5 c6xd5 Bd3-c2 g7-g5 Bd2-b4 g5-g4 Nf3-d2
h6-h5 Bb4xe7 Qd8xe7) -12.09/30 32} Nc6 {(Nb8-c6 Ng1-f3 Ng8-f6 Bf1-b5 g7-g6
Bb5xc6+ b7xc6 Nb1-c3) +8.92/8 163} 3. Nf3 {(Ng1-f3 a7-a6 a2-a3 e7-e6 c2-c4
g7-g6 c4xd5 e6xd5 Bf1-d3 Ng8-e7 Nb1-c3 Bf8-g7 O-O O-O Bd3-c2 f7-f6 Bc2-b3
Kg8-h8 Bc1-d2 Bc8-f5 Nf3-e1 Nc6-a5 Bb3-a2 Na5-c4 Ba2xc4 d5xc4 Ne1-f3 h7-h5
Rf1-d1 Kh8-g8 Rd1-c1) -11.91/33 59} Nf6 {(Ng8-f6 Bf1-b5 g7-g6 Bb5xc6+ b7xc6
Nb1-c3) +8.92/6 3} 4. Bb5 {(Bf1-b5 Nf6-d7 O-O Nc6-b8 Nf3-e5 c7-c6 Ne5xd7
Nb8xd7 Bb5-d3 e7-e6 c2-c4 Bf8-d6 Nb1-c3 O-O Bc1-d2 d5xc4 Bd3xc4 Nd7-f6
f2-f4 Bd6-e7 Ra1-d1 Bc8-d7 Rf1-f3 c6-c5 d4xc5 Be7xc5) -11.71/29 23} Bf5
{(Bc8-f5 Nf3-e5 Qd8-d6 Bb5xc6+ b7xc6 c2-c4) +8.87/6 1} 5. Ne5 {(Nf3-e5
Qd8-d6 Ne5xc6 b7xc6 Bb5-d3 Bf5-d7 b2-b3 Qd6-b4+ c2-c3 Qb4-a5 O-O c6-c5
Bc1-a3 c5xd4 e3xd4 e7-e6 Ba3xf8 Ke8xf8 Rf1-c1 c7-c5 d4xc5 Qa5xc5 h2-h3
Qc5-d6 a2-a3 e6-e5 Nb1-d2 Kf8-g8 Bd3-e2 a7-a5 Nd2-f3 a5-a4) -11.32/28 31
White resigns} *[/pgn]

[lkaufman]

a bit more info about chess handicaps
https://en.wikipedia.org/wiki/Handicap_(chess)
here's a bit more from the historical perspective:
http://www.edochess.ca/batgirl/odds-giving.html

Uri Blass claimed in this forum he can beat most
chess players with queen plus double rook odds.
(but then includes unrated players). well there are|
players who hardly know the rules, so it might be true.

As for ratings vs material odds, it's a better idea to look
at rating ratio (vs difference) so Nakamura vs 1400 is
2.0 ratio, but that's also for a 1200 vs 600 player. Queens odds?
maybe , but that is approx 8.5 points and to get a more
balanced game maybe other material odds also can be
applied (eg. rook plus bishop odds instead of queen).
A matter of trial and error, but then again, accuracy isn't
so important i guess, because i think of a self-correcting
system (of rating ratio's). More simple would be just
differences for rating 'classes' as superGM, GM, IM, NM
A,B,C etc. players

I just found out that Stockfish latest version can NOT beat TTE rated around 1480 with a Queen Odds at TC 60'+1" Probably the latest version of Komodo can, since Larry Kaufman gave special instructions to play with Odds, but it has to be tested

I have found that in a typical Elementary School chess club, only the top 10-15% are able to beat me with queen and two rooks handicap, the average such kid needs another knight or even two knights removed as well to have a chance. Usually the best player in the school needs about queen and knight odds, though of course occasionally some school has a kid who can win at queen odds.
Yes, it does make some sense to use the ratio of ratings for handicaps, although ratings are inherently linear, because the frequency and magnitude of errors declines as ratings rise, so knight odds for example equates to a much higher win percentage at GM level than at novice level.
I have lots of coverage of handicap chess, both between humans and in engine vs human games, in my soon-to-be published book "Chess Board Options". In my opinion, the most important question regarding handicap chess is which handicap or handicaps should fill the large gap between f7 odds and b1 (or g1) odds. The historical answer was "pawn and two moves", but even with this one included the gap to knight odds is still quite large, and even "pawn and three" still leaves a pretty sizable gap to knight odds.

[Chessqueen]

a bit more info about chess handicaps
https://en.wikipedia.org/wiki/Handicap_(chess)
here's a bit more from the historical perspective:
http://www.edochess.ca/batgirl/odds-giving.html

Uri Blass claimed in this forum he can beat most
chess players with queen plus double rook odds.
(but then includes unrated players). well there are|
players who hardly know the rules, so it might be true.

As for ratings vs material odds, it's a better idea to look
at rating ratio (vs difference) so Nakamura vs 1400 is
2.0 ratio, but that's also for a 1200 vs 600 player. Queens odds?
maybe , but that is approx 8.5 points and to get a more
balanced game maybe other material odds also can be
applied (eg. rook plus bishop odds instead of queen).
A matter of trial and error, but then again, accuracy isn't
so important i guess, because i think of a self-correcting
system (of rating ratio's). More simple would be just
differences for rating 'classes' as superGM, GM, IM, NM
A,B,C etc. players

I just found out that Stockfish latest version can NOT beat TTE rated around 1480 with a Queen Odds at TC 60'+1" Probably the latest version of Komodo can, since Larry Kaufman gave special instructions to play with Odds, but it has to be tested

I have found that in a typical Elementary School chess club, only the top 10-15% are able to beat me with queen and two rooks handicap, the average such kid needs another knight or even two knights removed as well to have a chance. Usually the best player in the school needs about queen and knight odds, though of course occasionally some school has a kid who can win at queen odds.
Yes, it does make some sense to use the ratio of ratings for handicaps, although ratings are inherently linear, because the frequency and magnitude of errors declines as ratings rise, so knight odds for example equates to a much higher win percentage at GM level than at novice level.
I have lots of coverage of handicap chess, both between humans and in engine vs human games, in my soon-to-be published book "Chess Board Options". In my opinion, the most important question regarding handicap chess is which handicap or handicaps should fill the large gap between f7 odds and b1 (or g1) odds. The historical answer was "pawn and two moves", but even with this one included the gap to knight odds is still quite large, and even "pawn and three" still leaves a pretty sizable gap to knight odds.

That is very good Larry, I just saw a video on Youtube where GM Nakamura gave a Queen Odds to a 1610 rated player and the 1610 player beat Nakamura very easy

Note: going back to the big difference between Stockfish and Komodo when giving Odds, Stockfish resigned against Tarrasch Toy rated 1480 and at the moment Komodo is even up the score against TTE, I did gave it a contemp of 180 to Komodo 13.3, I will post the result in 30 Minutes or 45 minutes, they are still playing

[Guenther]

I just found out that Stockfish latest version can NOT beat TTE rated around 1480 with a Queen Odds at TC 60'+1" Probably the latest version of Komodo can, since Larry Kaufman gave special instructions to play with Odds, but it has to be tested :roll:

[pgn][Event "Computer chess game"]
[Site "DESKTOP-OFQ3C0P"]
[Date "2021.02.11"]
[Round "?"]
[White "Stockfish_20090216_x64_modern"]
[Black "TarraschToyEngineV0.906"]
[Result "*"]
[BlackElo "1480"]
[Time "17:16:07"]
[WhiteElo "3480"]
[TimeControl "3600+5"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNB1KBNR w KQkq - 0 1"]
[Termination "normal"]
[PlyCount "9"]
[WhiteType "program"]
[BlackType "program"]

1. d4 {(d2-d4 e7-e6 Nb1-c3 Bf8-e7 Ng1-f3 d7-d6 e2-e4 Ng8-f6 Bc1-g5 Nb8-d7
e4-e5 d6xe5 Nf3xe5 Nd7xe5 d4xe5 Nf6-d7 Bg5xe7 Qd8xe7 O-O-O Nd7xe5 Rd1-e1
Ne5-g4 Re1-e2 O-O a2-a3 h7-h6 Nc3-e4) -12.24/31 110} d5 {(d7-d5 Nb1-c3
Nb8-c6 e2-e3 Ng8-f6 f2-f4 Bc8-f5 Bf1-d3) +8.90/8 60} 2. e3 {(e2-e3 e7-e6
Ng1-f3 c7-c6 c2-c4 Bf8-d6 Bf1-d3 Ng8-e7 Nb1-c3 Nb8-d7 O-O O-O Bc1-d2 a7-a6
Ra1-e1 Bd6-c7 Nc3-e2 h7-h6 c4xd5 c6xd5 Bd3-c2 g7-g5 Bd2-b4 g5-g4 Nf3-d2
h6-h5 Bb4xe7 Qd8xe7) -12.09/30 32} Nc6 {(Nb8-c6 Ng1-f3 Ng8-f6 Bf1-b5 g7-g6
Bb5xc6+ b7xc6 Nb1-c3) +8.92/8 163} 3. Nf3 {(Ng1-f3 a7-a6 a2-a3 e7-e6 c2-c4
g7-g6 c4xd5 e6xd5 Bf1-d3 Ng8-e7 Nb1-c3 Bf8-g7 O-O O-O Bd3-c2 f7-f6 Bc2-b3
Kg8-h8 Bc1-d2 Bc8-f5 Nf3-e1 Nc6-a5 Bb3-a2 Na5-c4 Ba2xc4 d5xc4 Ne1-f3 h7-h5
Rf1-d1 Kh8-g8 Rd1-c1) -11.91/33 59} Nf6 {(Ng8-f6 Bf1-b5 g7-g6 Bb5xc6+ b7xc6
Nb1-c3) +8.92/6 3} 4. Bb5 {(Bf1-b5 Nf6-d7 O-O Nc6-b8 Nf3-e5 c7-c6 Ne5xd7
Nb8xd7 Bb5-d3 e7-e6 c2-c4 Bf8-d6 Nb1-c3 O-O Bc1-d2 d5xc4 Bd3xc4 Nd7-f6
f2-f4 Bd6-e7 Ra1-d1 Bc8-d7 Rf1-f3 c6-c5 d4xc5 Be7xc5) -11.71/29 23} Bf5
{(Bc8-f5 Nf3-e5 Qd8-d6 Bb5xc6+ b7xc6 c2-c4) +8.87/6 1} 5. Ne5 {(Nf3-e5
Qd8-d6 Ne5xc6 b7xc6 Bb5-d3 Bf5-d7 b2-b3 Qd6-b4+ c2-c3 Qb4-a5 O-O c6-c5
Bc1-a3 c5xd4 e3xd4 e7-e6 Ba3xf8 Ke8xf8 Rf1-c1 c7-c5 d4xc5 Qa5xc5 h2-h3
Qc5-d6 a2-a3 e6-e5 Nb1-d2 Kf8-g8 Bd3-e2 a7-a5 Nd2-f3 a5-a4) -11.32/28 31
White resigns} *[/pgn]

Sigh, that's wrong! Your resign setting let the GUI adjudicate the game as won for TTE after 5 moves.
Actually SF thinks it has already 1 pawn of 12 won back...

[Chessqueen]

I just found out that Stockfish latest version can NOT beat TTE rated around 1480 with a Queen Odds at TC 60'+1" Probably the latest version of Komodo can, since Larry Kaufman gave special instructions to play with Odds, but it has to be tested


Sigh, that's wrong! Your resign setting let the GUI adjudicate the game as won for TTE after 5 moves.
Actually SF thinks it has already 1 pawn of 12 won back...

Thanks for mentioning that, I was going to ask that question,. How do I set up Arena so Stockfish does NOT resign. Actually I simply installed Arena about 11 months ago and I have NOT changed a single settings, but I would like to learn since it is in my believe that if I can adjust Arena to NOT tell Stockfish or NOT to resign for stockfish, eventually it will at least draw,I wonder is stockfish settings tell stockfish to resign when it sense that is down by -11.00, since I am now testing it against pigeon 1.5.1 rated 1799 with a Rook Odds and it is playing the entire game; but here with Komodo all that I had to do is to set up the contempt to 180 and it never resigned. Noticed how Komodo created an impenetrable fortress with pawns making the extra Queen useless

[pgn][Event "Computer chess game"]
[Site "DESKTOP-OFQ3C0P"]
[Date "2021.02.12"]
[Round "?"]
[White "Komodo-13.3-64bit"]
[Black "TarraschToyEngineV0.906"]
[Result "1/2-1/2"]
[BlackElo "1480"]
[Time "06:55:54"]
[WhiteElo "3480"]
[TimeControl "2700+5"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNB1KBNR w KQkq - 0 1"]
[Termination "normal"]
[PlyCount "73"]
[WhiteType "program"]
[BlackType "program"]

1. Nf3 d5 2. c3 Nc6 3. d4 Nf6 4. Bf4 g6 5. g3 Bg4 6. Nbd2 Bg7 7. Bg2 Bxf3
8. Nxf3 Qd7 9. O-O O-O 10. Ne1 Ne4 11. Nd3 b6 12. Rad1 f5 13. f3 Nd6 14.
Rfe1 a5 15. e3 a4 16. Bxd6 cxd6 17. a3 e6 18. Kf2 e5 19. f4 e4 20. Nc1 Na5
21. Bf1 Rfc8 22. Na2 Qf7 23. h4 Bf6 24. Be2 Nc4 25. Bxc4 Rxc4 26. Rh1 Rcc8
27. Rh2 Kg7 28. Rdh1 Qe7 29. Nb4 Qf7 30. Rg2 Rab8 31. Kg1 Ra8 32. Rf2 Rab8
33. Kg2 Ra8 34. Nc2 Rab8 35. Nb4 Rb7 36. Na2 Rbb8 37. Nb4 {3-fold
repetition} 1/2-1/2[/pgn]

[jefk]

chess in Father style

any human, also rated around 1450 would try
to break through that fortress !

Also wondering if there are no other engines (or an engine
that you can set at approx 1450 level) which wouldnt
play as stupid as this TTE, but then again,
at rating ratio around two, i would prefer a handicap
with eg. rook, bishop, and knight and pawn or so
rather than the mighty queen...


PS some historical games with queen odds
(only Morphy lost
https://www.chessgames.com/perl/chessco ... id=1009181

[Chessqueen]

chess in Father style

any human, also rated around 1450 would try
to break through that fortress !

Also wondering if there are no other engines (or an engine
that you can set at approx 1450 level) which wouldnt
play as stupid as this TTE, but then again,
at rating ratio around two, i would prefer a handicap
with eg. rook, bishop, and knight and pawn or so
rather than the mighty queen...


PS some historical games with queen odds
(only Morphy lost
https://www.chessgames.com/perl/chessco ... id=1009181

But a handicap of Rook + Bishop + Knight alone without any pawn is greater than the Queen

[pgn][Event "Computer chess game"]
[Site "DESKTOP-OFQ3C0P"]
[Date "2021.02.12"]
[Round "?"]
[White "Stockfish_20090216_x64_modern"]
[Black "Pigeon-1.5.1"]
[Result "1/2-1/2"]
[BlackElo "1799"]
[Time "08:25:12"]
[WhiteElo "3490"]
[TimeControl "900+3"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/1NBQKBNR w Kkq - 0 1"]
[Termination "normal"]
[PlyCount "56"]
[WhiteType "program"]
[BlackType "program"]

1. e4 {(e2-e4 e7-e6 d2-d4 d7-d5 Nb1-d2 c7-c5 Ng1-f3 Ng8-f6 e4xd5 e6xd5
Bf1-b5+ Bc8-d7 Bb5xd7+ Nb8xd7 O-O Bf8-e7 d4xc5 Nd7xc5 Nf3-d4 O-O Nd4-f5
Qd8-d7 Qd1-f3 Be7-d6 Nd2-b3) -6.15/26 47} e6 {(e7-e6 Nb1-c3 Nb8-c6 Ng1-f3
Ng8-f6 d2-d4 Nc6-b4 Nf3-e5 Bf8-d6) +4.62/9 39} 2. d4 {(d2-d4 d7-d5 Nb1-c3
Ng8-f6 e4-e5 Nf6-d7 Nc3-e2 c7-c5 c2-c3 Nb8-c6 Ng1-f3 Qd8-c7 a2-a3 Bf8-e7
g2-g3 b7-b6 Bf1-g2 Bc8-a6 O-O O-O Qd1-c2 Ba6-c4 Rf1-e1 a7-a5 b2-b3 Bc4xe2
Re1xe2 b6-b5 Kg1-h1 a5-a4) -6.17/26 41} Nf6 {(Ng8-f6 e4-e5 Nf6-d5 c2-c4
Bf8-b4+ Bc1-d2 Bb4xd2+ Nb1xd2) +4.39/8 15} 3. e5 {(e4-e5 Nf6-d5 c2-c4
Nd5-b6 Nb1-c3 d7-d6 Ng1-f3 Nb8-c6 e5xd6 c7xd6 Bc1-e3 d6-d5 c4-c5 Nb6-d7
a2-a3 e6-e5 d4xe5 Nd7xc5 Bf1-b5 a7-a6 Bb5xc6+ b7xc6) -5.48/23 7} Ne4
{(Nf6-e4 Bf1-d3 d7-d5 e5xd6/ep Ne4xd6 Ng1-f3 Nb8-c6 Bc1-g5) +4.21/8 45} 4.
Ne2 {(Ng1-e2 f7-f5 Ne2-f4 g7-g6 Bf1-b5 c7-c6 Bb5-d3 d7-d6 O-O d6xe5 d4xe5
Ne4-c5 Nb1-d2 Nc5xd3 Nf4xd3 b7-b6 b2-b3 Bc8-a6 Nd2-c4 Ba6xc4 b3xc4 a7-a5
Nd3-f4 Qd8xd1 Rf1xd1 Ke8-f7) -5.57/26 86} Qh4 {(Qd8-h4 g2-g3 Qh4-g4 Nb1-d2
Bf8-b4 c2-c3 Qg4-f5 Ne2-f4) +4.69/8 12} 5. g3 {(g2-g3 Qh4-d8 Bf1-g2 f7-f5
h2-h4 c7-c5 f2-f3 Ne4xg3 Ne2xg3 c5xd4 f3-f4 Nb8-c6 Ng3-e2 Qd8-a5+ Nb1-d2
Bf8-c5 a2-a3 Bc5-b6 b2-b4 Qa5-a6 Nd2-b3 O-O Ne2xd4 Nc6xd4 Nb3xd4 Qa6-c4)
-3.97/21 5} Qg4 {(Qh4-g4 Bf1-g2 Qg4-f5 f2-f3 Ne4-g5 Qd1-d3 Qf5xd3 c2xd3)
+4.26/7 2} 6. Bg2 {(Bf1-g2 Qg4-f5 f2-f3 Ne4-c5 g3-g4 Qf5-g6 d4xc5 Bf8xc5
Nb1-c3 f7-f5 e5xf6/ep Qg6xf6 Nc3-e4 Qf6-h4+ Ne2-g3 Bc5-e7 c2-c4 O-O O-O
Nb8-c6 a2-a3 b7-b6 f3-f4 Bc8-b7) -2.75/22 9} Qf5 {(Qg4-f5 O-O h7-h6 Qd1-d3
Ne4-g5 Qd3xf5 e6xf5) +3.82/7 6} 7. f3 {(f2-f3 Ne4-c5 g3-g4 Qf5-g6 h2-h4
f7-f5 d4xc5 Bf8xc5 Ne2-f4 Qg6-f7 Nb1-c3 Nb8-c6 Nf4-d3 Bc5-d4 Nc3-b5 Bd4-b6
h4-h5 a7-a6 Nb5-c3 d7-d5 e5xd6/ep e6-e5 Nd3-f2 Bc8-e6 g4xf5 Be6xf5 d6xc7)
-3.11/24 20} Ng5 {(Ne4-g5 g3-g4 Ng5xf3+ Ke1-f2 Qf5xg4 Bg2xf3 Qg4-h4+
Kf2-g1) +3.27/8 14} 8. Nf4 {(Ne2-f4 h7-h5 h2-h4 Ng5xf3+ Bg2xf3 d7-d6 e5xd6
Bf8xd6 O-O Nb8-c6 Nf4xh5 Qf5-g6 Bc1-g5 e6-e5 c2-c3 O-O Nb1-d2 e5xd4 Bf3-e4
f7-f5 Be4-b1 d4xc3 b2xc3 Qg6-e6 Nd2-f3 Qe6-c4 Kg1-g2 Kg8-h8) -2.64/22 12}
h6 {(h7-h6 g3-g4 Bf8-b4+ Ke1-f1 Qf5-h7 h2-h4 f7-f6 h4xg5 f6xg5) +2.66/8 42}
9. h4 {(h2-h4 d7-d5 g3-g4 Ng5xf3+ Bg2xf3 Qf5-h7 Bc1-e3 Nb8-c6 c2-c4 d5xc4
Nb1-d2 Bc8-d7 O-O O-O-O Bf3-e4 Qh7-g8 Nd2xc4 Kc8-b8 h4-h5 Bf8-e7 Qd1-b3
Be7-g5 Rf1-c1 Bg5xf4 Be3xf4) -2.28/24 34} Bb4+ {(Bf8-b4+ Nb1-c3 Nb8-c6
g3-g4 Ng5xf3+ Bg2xf3 Qf5-h7 g4-g5) +2.35/8 11} 10. Nc3 {(Nb1-c3 d7-d5
Nf4-h5 Ng5xf3+ Bg2xf3 Qf5-h7 O-O Nb8-c6 Bf3-e2 Bc8-d7 Rf1-f4 Bb4-e7 Nc3-b5
O-O-O Rf4xf7 Be7-f8 Kg1-h2 Qh7-g8 Rf7-f2 g7-g5 Nb5-c3 g5xh4 g3xh4) -1.67/20
13} Nc6 {(Nb8-c6 g3-g4 Ng5xf3+ Bg2xf3 Qf5-h7 a2-a3 Bb4-e7) +2.37/7 3} 11.
hxg5 {(h4xg5 Qf5xg5 g3-g4 Qg5-d8 Nf4-h5 d7-d6 f3-f4 Rh8-g8 Qd1-d3 Bc8-d7
Qd3-h7 Ke8-e7 Nh5xg7 Qd8-f8 Ng7-h5 Nc6xd4 Nh5-f6 Qf8-g7 Rh1xh6 Qg7xh7
Rh6xh7 d6xe5) -1.17/28 47} Qxg5 {(Qf5xg5 Ke1-f2 Qg5-e7 Bc1-e3 a7-a6 Nf4-d3
Rh8-h7 d4-d5) +2.20/8 14} 12. Kf2 {(Ke1-f2 Qg5-d8 Nf4-h5 Ke8-f8 Bc1-e3
Bb4-e7 d4-d5 e6xd5 Nc3xd5 d7-d6 e5xd6 Be7xd6 f3-f4 Bc8-e6 Bg2-f3 Qd8-d7
g3-g4 f7-f5 g4xf5 Be6xf5 Nh5-g3 Bf5-g6 c2-c4 Ra8-d8) -1.69/26 25} Qe7
{(Qg5-e7 Bc1-e3 b7-b6 Nf4-d3 Bc8-a6 d4-d5 Ba6xd3 d5xc6) +2.30/7 5} 13. Nh5
{(Nf4-h5 Qe7-f8 Nc3-b5 Bb4-a5 f3-f4 a7-a6 Nb5-a3 b7-b5 c2-c4 Bc8-b7 Bc1-e3
Ba5-b6 c4-c5 Bb6-a7 Na3-c2 O-O-O a2-a4 b5xa4 Qd1-a1 Nc6-e7 Bg2xb7+ Kc8xb7
Qa1xa4 c7-c6 Kf2-g1 Ne7-d5 Kg1-g2) -1.69/26 55} Kf8 {(Ke8-f8 Bc1-e3 Bb4xc3
b2xc3 Qe7-a3 Qd1-a1 Rh8-h7 Rh1-h3) +2.31/8 18} 14. Ne2 {(Nc3-e2 Qe7-d8
g3-g4 Bb4-e7 f3-f4 g7-g6 Nh5-g3 Be7-h4 Bg2-f3 d7-d5 c2-c3 Kf8-g7 Kf2-g2
a7-a5 Bc1-d2 a5-a4 Bd2-e3 Nc6-a5 Qd1xa4 Bc8-d7 Qa4-c2) -1.19/22 5} d6
{(d7-d6 c2-c3 Bb4-a5 f3-f4 d6xe5 f4xe5 Bc8-d7 Qd1-d3) +2.48/8 17} 15. f4
{(f3-f4 Bb4-a5 a2-a3 Ba5-b6 c2-c4 g7-g6 Bg2xc6 b7xc6 Nh5-f6 h6-h5 g3-g4
h5-h4 b2-b4 c6-c5 b4xc5 d6xc5 d4-d5 e6xd5 c4xd5 c5-c4+ Kf2-g2 Kf8-g7 Ne2-c3
Qe7-c5 Nc3-e4 Qc5-d4 Qd1-f3) -0.89/23 25} dxe5 {(d6xe5 f4xe5 Bc8-d7 c2-c3
Bb4-a5 Bc1-e3 Rh8-h7 Bg2-e4) +2.29/8 9} 16. c3 {(c2-c3 e5xf4 c3xb4 f4xg3+
Ne2xg3 Qe7xb4 Bg2xc6 b7xc6 a2-a3 Qb4-e7 Ng3-e4 e6-e5 d4xe5 Bc8-f5 Ne4-g3
Bf5-g6 Qd1-g4 Qe7-c5+ Bc1-e3 Qc5-c2+ Ng3-e2 Ra8-e8 Be3-d4 Qc2-e4 Qg4-h4
Qe4xh4+ Rh1xh4 Kf8-g8) -0.29/23 21} exf4 {(e5xf4 c3xb4 f4xg3+ Kf2xg3 Qe7xb4
Qd1-c2 Nc6xd4 Ne2xd4 Qb4xd4 Qc2xc7) +2.55/9 22} 17. cxb4 {(c3xb4 f4xg3+
Ne2xg3 Qe7xb4 Bg2xc6 b7xc6 Nh5xg7 Kf8xg7 Rh1xh6 Rh8xh6 Bc1xh6+ Kg7xh6
Qd1-h5+ Kh6-g7 Qh5-g5+ Kg7-f8 Qg5-d8+ Kf8-g7) 0.00/25 7} fxg3+ {(f4xg3+
Kf2xg3 Qe7xb4 Qd1-b3 Qb4xb3+ a2xb3 g7-g5 Nh5-f6) +2.40/8 6} 18. Nexg3
{(Ne2xg3 Qe7xb4 Bg2xc6 b7xc6 Nh5xg7 Kf8xg7 Rh1xh6 Rh8xh6 Bc1xh6+ Kg7xh6
Qd1-h5+ Kh6-g7 Qh5-g5+ Kg7-f8 Qg5-d8+ Kf8-g7) 0.00/26 10} Qxb4 {(Qe7xb4
Ng3-e2 g7-g6 Nh5-f4 Qb4-c4 b2-b3 Qc4-b4 Bg2-e4) +2.56/8 14} 19. Nxg7
{(Nh5xg7 Kf8xg7 Bg2xc6 b7xc6 Rh1xh6 Rh8xh6 Bc1xh6+ Kg7xh6 Qd1-h5+ Kh6-g7
Qh5-g5+ Kg7-h7 Qg5-h5+) 0.00/25 10} Kxg7 {(Kf8xg7 Bg2xc6 b7xc6 Qd1-g4+
Kg7-f8 Rh1xh6 Rh8xh6 Bc1xh6+ Kf8-e7) +3.67/9 30} 20. Bxc6 {(Bg2xc6 b7xc6
Rh1xh6 Rh8xh6 Bc1xh6+ Kg7xh6 Qd1-h5+ Kh6-g7 Qh5-g5+ Kg7-h7 Qg5-h5+) 0.00/33
9} bxc6 {(b7xc6 a2-a3 Qb4-b5 Qd1-g4+ Kg7-f6 Qg4-f4+ Kf6-e7 Qf4xc7+ Bc8-d7)
+3.53/8 5} 21. Rxh6 {(Rh1xh6 Rh8xh6 Bc1xh6+ Kg7xh6 Qd1-h5+ Kh6-g7 Qh5-g5+
Kg7-h7 Qg5-h5+) 0.00/36 11} Rxh6 {(Rh8xh6 Bc1xh6+ Kg7xh6 Qd1-h5+ Kh6-g7
Qh5-e5+ Kg7-g8 Qe5-g5+ Kg8-f8) +6.62/9 12} 22. Bxh6+ {(Bc1xh6+ Kg7xh6
Qd1-h5+ Kh6-g7 Qh5-g5+ Kg7-h7 Qg5-h5+) 0.00/41 18} Kxh6 {(Kg7xh6 Qd1-c1+
Kh6-g6 Ng3-e2 Bc8-d7 a2-a3 Qb4-b3 Qc1-c3 Ra8-b8 Ne2-f4+ Kg6-f5) +6.50/10
15} 23. Qh5+ {(Qd1-h5+ Kh6-g7 Qh5-g5+ Kg7-h7 Qg5-h5+ Kh7-g7) 0.00/53 4} Kg7
{(Kh6-g7 Qh5-g5+ Kg7-f8 Qg5-d8+ Kf8-g7 Qd8-g5+ Kg7-h7 Qg5-h4+ Kh7-g6
Qh4-h5+ Kg6-g7) +6.69/11 11} 24. Qg5+ {(Qh5-g5+ Kg7-h7 Qg5-h5+ Kh7-g8
Qh5-g5+) 0.00/50 9} Kh7 {(Kg7-h7 Ng3-e2 Qb4xb2 Qg5-f6 Kh7-g8 Qf6-g5+ Kg8-h7
Qg5-e7 Kh7-g8 Qe7xc7) +6.72/10 10} 25. Qh5+ {(Qg5-h5+ Kh7-g8 Qh5-g5+ Kg8-h8
Qg5-h4+ Kh8-g8) 0.00/55 16} Kg8 {(Kh7-g8 Qh5-g5+ Kg8-f8 Qg5-d8+ Kf8-g7
Qd8-g5+ Kg7-h7 Qg5-h5+ Kh7-g8 Qh5-g5+ Kg8-f8) +6.61/11 5} 26. Qg5+
{(Qh5-g5+ Kg8-h7 Qg5-h5+ Kh7-g8) 0.00/52 4} Kh7 {(Kg8-h7 Qg5-h4+ Kh7-g7
Qh4-g5+ Kg7-f8 Ng3-e2 Bc8-a6 Qg5-e5 Kf8-g8 b2-b3 Qb4-d2) +6.69/11 16} 27.
Qh5+ {(Qg5-h5+ Kh7-g7 Qh5-g5+) 0.00/53 6} Kg8 {(Kh7-g8 Qh5-g5+ Kg8-f8
Qg5-d8+ Kf8-g7 Qd8-g5+ Kg7-f8 Ng3-e2 Bc8-a6 Qg5-c5+ Qb4xc5 d4xc5) +6.62/12
15} 28. Qg5+ {(Qh5-g5+ Kg8-f8 Qg5-d8+ Kf8-g7 Qd8-g5+) 0.00/55 32} Kh7
{(Kg8-h7 Qg5-h4+ Kh7-g7 Qh4-g5+ Kg7-h7 Qg5-h4+ Kh7-g7 Qh4-g5+ Kg7-h7
Qg5-h5+ Kh7-g8) +6.73/10 6 3-fold repetition} 1/2-1/2[/pgn]

[jefk]

CQ :

But a handicap of Rook + Bishop + Knight alone without any pawn is greater than the Queen

ok, yep, then you can give a pawn handicap to the other side

Personally I'm not so interested in such big material odds,
although it could happen in a diverse internal chess-club
competition between a newcomer/beginner of eg 800
vs someone around 1600.
Also, this thread is about a rook handicap, isn't it ?


PS in the Fair-Chess FC system that i'm thinking of, there
will be various classes for different rating ratio's
from e.g 1-1.3 1.3-1.6 1.6-1.9 and higher than 1.9
Only in the latter we'll probably need a higher handicap
than just a rook... my2cts

Don't know why you are interested in handicap chess, but anyway
in the thread about correspondence chess i brought up why i'm
seriously thinking about such a variant. How will it work ?
Well lets say via the computerized weekly interntal competition
in a chess club only a few people show up, of very different
rating categories, and then i as 1700 or so have to play
against a unrated (we estimate 900) player. Even although
we have also setup some comp to play ..
So rating ratio (>) 1.9 Then boom, i get the big handicap,
no queen (or rook/bishop/kn) whatever, and bang,
my rating drops (it becomes a temporary 1700/1.9) and
after the game, the new ratings are calculated as result
of the game in two steps, first as Elo, eg. if i lose, then
newcomer becomes eg. 930, and i will become eg 880 (depending
on how much games i played earlier), but then then multiplied
again with 1.9 , so that will be (trala... 1672, believe it or not,
and of course it would be only an indicative internal rating
within the chess club, not like the Fide rating (and yes the late
mr Elo will probably will get a heart attack in his afterlife
but that's not my concern). So it works ! Maybe i've should
have called it Fun-Chess (instead of Fair-ches), but who cares,
and there already this name... :
https://www.chessvariants.com/other.dir ... chess.html
And it's not about the name, but about the method (crazyhouse
is not crazy but fun as well, although not many chess players
seem to know it (yet) ; wait until Fair-Crazy-House chess i suggest)


(*)still working on it, and LK probably will come
with many suggestions

[Chessqueen]

CQ :

But a handicap of Rook + Bishop + Knight alone without any pawn is greater than the Queen

ok, yep, then you can give a pawn handicap to the other side

Personally I'm not so interested in such big material odds,
although it could happen in a diverse internal chess-club
competition between a newcomer/beginner of eg 800
vs someone around 1600.
Also, this thread is about a rook handicap, isn't it ?


Yes it is about a Rook Odds, you are correct, and the reason why I became interested in Odds is because I saw GM Nakamura giving a Queen Odds to a 1638+ rated player and I was also wondering if Stockfish which is rated 3490 or Komodo 3470+ can beat a 1650 thru 1700 human player with a Queen Odds, so I picked Piranha rated around 1699 and I just found out that even with a Knight Odds plus two pawns Stockfish latest version is having a hard time to beat Piranha, but humans get lots of pressure whereas engine do NOT

Here is the video of GM Nakamura losing to a 1638 with a Queen Odds


Piranha with a little bit more time like 5 more minutes could have beaten Stockfish

[pgn][Event "Computer chess game"]
[Site "DESKTOP-OFQ3C0P"]
[Date "2021.02.12"]
[Round "?"]
[White "Stockfish_20090216_x64_modern"]
[Black "Piranha rated 1699"]
[Result "1/2-1/2"]
[BlackElo "1699"]
[Time "10:30:44"]
[WhiteElo "3490"]
[TimeControl "900+3"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/1P1PPPPP/R1BQKBNR w KQkq - 0 1"]
[Termination "normal"]
[PlyCount "75"]
[WhiteType "program"]
[BlackType "program"]

1. Nf3 Nc6 2. d4 e6 3. e4 d5 4. e5 Nh6 5. Bxh6 gxh6 6. Bb5 Bb4+ 7. Kf1 Bd7
8. Bxc6 Bxc6 9. g3 O-O 10. h4 f6 11. h5 fxe5 12. Nxe5 Bb5+ 13. Kg2 Qg5 14.
Rh4 Qd2 15. Qxd2 Bxd2 16. Rhh1 Be2 17. f4 Rf5 18. Ra2 Bxh5 19. Ra3 Kh8 20.
Rb3 Rg8 21. Kf2 b6 22. Nc6 Rgf8 23. Rd3 Ba5 24. b4 Be8 25. Nxa5 bxa5 26.
bxa5 R5f6 27. Rc3 c6 28. Rb3 Rg8 29. Rb7 Rg7 30. a6 h5 31. Rb8 Rg8 32. Ra8
Rfg6 33. Rxh5 Rxg3 34. Rxe8 Rg2+ 35. Kf3 Rg3+ 36. Kf2 Rg2+ 37. Kf3 Rg3+ 38.
Kf2 {3-fold repetition} 1/2-1/2[/pgn]

[Uri Blass]

a bit more info about chess handicaps
https://en.wikipedia.org/wiki/Handicap_(chess)
here's a bit more from the historical perspective:
http://www.edochess.ca/batgirl/odds-giving.html

Uri Blass claimed in this forum he can beat most
chess players with queen plus double rook odds.
(but then includes unrated players). well there are|
players who hardly know the rules, so it might be true.

As for ratings vs material odds, it's a better idea to look
at rating ratio (vs difference) so Nakamura vs 1400 is
2.0 ratio, but that's also for a 1200 vs 600 player. Queens odds?
maybe , but that is approx 8.5 points and to get a more
balanced game maybe other material odds also can be
applied (eg. rook plus bishop odds instead of queen).
A matter of trial and error, but then again, accuracy isn't
so important i guess, because i think of a self-correcting
system (of rating ratio's). More simple would be just
differences for rating 'classes' as superGM, GM, IM, NM
A,B,C etc. players

I just found out that Stockfish latest version can NOT beat TTE rated around 1480 with a Queen Odds at TC 60'+1" Probably the latest version of Komodo can, since Larry Kaufman gave special instructions to play with Odds, but it has to be tested

I have found that in a typical Elementary School chess club, only the top 10-15% are able to beat me with queen and two rooks handicap, the average such kid needs another knight or even two knights removed as well to have a chance. Usually the best player in the school needs about queen and knight odds, though of course occasionally some school has a kid who can win at queen odds.
Yes, it does make some sense to use the ratio of ratings for handicaps, although ratings are inherently linear, because the frequency and magnitude of errors declines as ratings rise, so knight odds for example equates to a much higher win percentage at GM level than at novice level.
I have lots of coverage of handicap chess, both between humans and in engine vs human games, in my soon-to-be published book "Chess Board Options". In my opinion, the most important question regarding handicap chess is which handicap or handicaps should fill the large gap between f7 odds and b1 (or g1) odds. The historical answer was "pawn and two moves", but even with this one included the gap to knight odds is still quite large, and even "pawn and three" still leaves a pretty sizable gap to knight odds.

1)What are the ages in typical elementary school chess club?
I know that in israel elementary school is usually between ages 7-12 when 7 is first grade and 12 is 6th grade.

2)I wonder about the meaning of the top 10-15% that are able to beat you with queen and 2 rooks handicap.
The top 10-15% of what group do you mean?

The top 10-15% of children who play in some chess club that you visit may be different than the top 10-15% of children who know the chess rules and
may be different than the top 10-15% of the children that part of them even do not know the chess rules.