Talk:Handicap (chess)

Historical Question
Just wondering ...

In capped pawn or ringed piece odds, did the designated chessman have to deliver the mating check directly, or was a discovered check uncovered by the move of the designated chessman sufficient? The legality of this alternative should make the odds slightly less overepowering, though notr much, I'd suspect. Also, it would permit a ringed king who could only deliver mate via discovery. WHPratt (talk) 14:27, 6 May 2014 (UTC)

Castling rights
Have castling rights (odd-giver forfeits right to castle to either side) ever bee a significant part of odds? It may seem a good thing in practice mode or coaching mode or to teach advantage and disadvantages of castling. Thieh (talk) 18:43, 28 May 2014 (UTC)


 * It'd be most interesting to test that via a computer chess engine, to determine what the disadvantage amounts to over a significant number of games between equal "players." Someone probably already has eveluated it. I'd guess a good bit less than a pawn, but just how much? WHPratt (talk) 17:21, 31 May 2014 (UTC)
 * It's probably a little bit more valuable than pawn promotion to knight. Just half-kidding. A very rough "estimate" perhaps can be achieved "instantly" with a custom FEN setting in almost any computer chess program, the "custom" aspect being only "no castling rights" to one side, and then looking just at the estimated advantage even before the first move. I even feel like installing one just to check this useless piece of knowledge. The value will likely change for black and white. A more "accurate" assessment would possibly require some less trivial hacking on the engines, so that the adversary "knows" of this handicap, and could exploit it, sure that the opponent won't castle. Out of this situation it probably will vary too much to be anything other than a meaningless value. Sometimes castling is the best move, sometimes it's a blunder, unlike a piece or a pawn, whose values are far less variable.
 * Apparently denying White's castling rights exactly nullifies the opening advantage, according to Stockfish 14.1. But Lc0 sees a disadvantage for White already. Double sharp (talk) 10:02, 9 December 2021 (UTC)

This kind of thing is discussed in Kaufman's Chess Board Options. I'll see if I can work it in. Double sharp (talk) 05:56, 20 September 2023 (UTC)

Nakamura, 2014
Nakamura played a series of games against a handicaped (one pawn less) Stockfish, but arguably with his own handicap of being human and playing for 10 hours straight, even though with the aid/compensation of an older version of Rybka, ~200 ELO points inferior to Stockfish. "Stockfish beats Nakamura". — Preceding unsigned comment added by 179.99.178.197 (talk) 04:43, 29 August 2014 (UTC)

"From this perspective, a game beginning from a "lost" position becomes less interesting, even distasteful."
I doubt pawn-and-move is a lost position for Black objectively. One can certainly hang on a pawn down and escape with a draw. Besides, though Black's king is exposed (the f-pawn is missing), (s)he can make a virtue out of this by quickly castling kingside and using the king's rook on the already half-open f-file. I would expect that with perfect play, Black could salvage a draw out of pawn-and-move.

Knight odds I am basically certain is lost for Black objectively, and I think pawn-and-two-moves may be so if White begins 1.e4 2.d4○, grabbing tons of space, central control, and preparing for lots of development. But this complaint doesn't really work for pawn-and-move, I think. Double sharp (talk) 16:10, 7 February 2016 (UTC)


 * Oh, I notice there are lower handicaps. I would surely call odds of the move not really odds at all, as any series of games between two players will have each player getting as many Blacks as Whites, and the initial position is certainly =. Two moves is probably +/= if White begins 1.e4 2.d4○. Pawn and move I would class as +/−, and the higher handicaps +− (White wins objectively). But the quoted complaint cannot apply to pawn and move and below, I think. Double sharp (talk) 16:15, 7 February 2016 (UTC)
 * By the way, how can exchange odds be a higher level than rook and knight? Double sharp (talk) 16:21, 7 February 2016 (UTC)

Perhaps a better reason to be against odds is the fact that they let you simply try to force as many exchanges as possible, until the stronger side has no pieces left. Double sharp (talk) 16:30, 7 February 2016 (UTC)

Has anyone managed to suggest some kind of odds-giving that would effectively neutralize the advantage of White having the first move, more-or-less exactly? It seems to me that such a concept could be very popular. Not that it would replace the conventional game, but might be employed for the occasional tiebreak. I know that they now use something with varying times on the clock, but I'm talking about something positional and very subtle. Say that Black begins with both rook pawns on the third rank, or say that White cannot play a Pawn double-step on the first move (only). Now, even those ideas are probably insufficient or too extreme, but, use your imagination! WHPratt (talk) 19:53, 14 May 2019 (UTC)
 * The last suggestion probably won't work, as 1.Nf3 and 1.g3 should not harm White's first-move advantage too badly judging from the stats at chessgames.com. Double sharp (talk) 12:24, 7 February 2020 (UTC)
 * Stockfish 14.1 claims that your proposed a6/h6 handicap is too large and provides -0.30 advantage to Black. Admittedly analysing opening positions with engines is iffy, but intuitively that seems plausible, since so many forceful openings for White require getting a bishop (or maybe a knight) to N5. Or, to adapt a quote by Botvinnik, "the Najdorf is so strong that you cannot let your opponent play it with an extra tempo". (Really he said it about the Sicilian, to explain why he thought it was bad to answer 1.c4 with 1...e5.) :D Double sharp (talk) 11:28, 15 November 2021 (UTC)

Stockfish 14.1 says: Two moves 1.e4 2.d4 (and then Black to move) is +0.94, which looks pretty grim. But Kaufman in Chess Board Options says it's really about +0.6 and not won yet, backing the latter statement up with empirical results. Pawn and move (f7) is +2.82, so I was completely wrong. And pawn (f2) is −1.56. Double sharp (talk) 09:55, 15 December 2021 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Handicap (chess). Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20071026233735/http://www.bobby-fischer.net/bobby_fischer_quotes_96.htm to http://www.bobby-fischer.net/bobby_fischer_quotes_96.htm
 * Added tag to http://home.comcast.net/~danheisman/Articles/evaluation_of_material_imbalance.htm

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 22:38, 28 October 2017 (UTC)

Pawn and four moves
Not explicitly stated by Kaufman (whose assessment of this handicap I added to the article), but using the brain prosthesis widely known as Stockfish 11 to supplement my poor skills: 1.e3 2.Bd3 3.Qg4 4.Nc3 is the problem with pawn and four moves handicap. Now White threatens mate in 2 by 5.Qg6+, and there are not many moves that defend against that:
 * 4...Nh6 is met by 5.Bxh7! Nf7 6.Bg6 e5 and Black is left with a rotten position with that awkward knight on f7.
 * 4...d6 invites 5.Qh5+ Kd7 6.Bf5+ e6 7.Qf7+ Qe7 8.Bxe6+ Kd8 9.Qf3 Qxe6 10.Qxf8+ Qe8 11.Qxg7 Ne7 and Black again has a rotten position, three pawns down: I guess this is the line that Kaufman considers.
 * 4...e6 may provide the best practical chances, although that is my WP:OR with a brain prosthesis (as Tim Krabbé has memorably put it). After 5.Qh5+ Ke7 6.Nf3 Nf6 7.Qh4 d6 8.e4 Nc6 9.Bc4 Ke8 10.0-0 Be7 11.d4, Black can't castle (which would otherwise look like a great move at the moment), but his position at least looks less like a trainwreck than in the other two lines, and given the odds that must be a psychological advantage against a White player who went straight for the kill with these not-quite-normal opening moves rather than something principled like 1.e4 2.d4 3.Nf3 4.Bc4. ^_^ Double sharp (talk) 22:39, 24 January 2020 (UTC)

Why the f-pawn?
If the stronger player remove one of his pawn, but the removed pawn is the a-pawn, or the b-pawn, or the c-pawn, …, what will be the result? 211.23.210.36 (talk) 21:56, 7 October 2021 (UTC)


 * I believe that losing any other pawn might give back some of the odds advantage to the stronger player, permitting him or her to develop an attack faster than in a conventional game. Removing other pawns opens up diagonals and more useful files. The f-pawn was probably designated for odds-giving to give the biggest advantage to the weaker side. The f-pawn is almost useless on offense, but important for defense (its absence exposes the king). WHPratt (talk) 03:37, 8 October 2021 (UTC)
 * is completely correct. Larry Kaufman remarks in Chess Board Options: A Memoir of Players, Games and Engines: "The point is that the removal of any other pawn offers at least some compensation in the form of an immediately useful open file for the rook or diagonal for bishop or queen." Double sharp (talk) 11:15, 15 November 2021 (UTC)

Stockfish 14.1 says: Double sharp (talk) 11:11, 15 December 2021 (UTC)
 * Remove a7: +1.67
 * Remove b7: +2.20
 * Remove c7: +2.06
 * Remove d7: +2.09
 * Remove e7: +2.39
 * Remove f7: +2.82
 * Remove g7: +2.54
 * Remove h7: +1.52

And: Double sharp (talk) 11:13, 15 December 2021 (UTC)
 * Remove a2: −0.82
 * Remove b2: −1.54
 * Remove c2: −1.54
 * Remove d2: −1.25
 * Remove e2: −1.16
 * Remove f2: −1.56
 * Remove g2: −1.66
 * Remove h2: −0.49

Anecdote
I'm not putting this anecdote on this article's front page because I don't have a source.

Alekhine, in his latter days, shabbily clad, was studying some chess game on a portable set while waiting for his train somewhere in Spain or Portugal. Another passenger arrives: “Oh, you're a chess player; could we play together? — Certainly”, says Alekhine, and he sets up the initial position without the Qd1. “B-b-but, you don't know me, and you're giving me queen odds? — If I couldn't”, says Alekhine, “I would know you.”

Tonymec (talk) 17:56, 20 January 2022 (UTC)

Other handicaps
In case anyone wondered: knight, pawn, and move (f7) is about the same as rook odds, per a forum post by Kaufman. Double sharp (talk) 04:17, 25 February 2024 (UTC)