Chess Handicaps

Jun 15, 2008, 3:54 AM |

Chess handicaps are an interest of mine. These handicaps are designed to provide a more spirited, equitable game between players of disparate strength. The historical origin of the chess handicap is that games were frequently played for stakes, with the weaker player refusing to take the game without a suitable handicap. There are several types of chess handicaps, some compatible with orthodox chess and some incompatible.

Common Handicaps Incompatible with Orthodox Chess

 Some handicaps are incompatible with orthodox chess. These include:

* Cheapo moves. Once per game, when announcing "Cheapo!" the weaker player is permitted to make two moves in succession. One supposes that the king must be removed from check in the first of the two moves if it started in check, but this is not always the case. In some systems, positions that are otherwise checkmate can be averted with a "Cheapo!" move.

 * Extra drop pieces. At any time, the weaker player is allowed to drop a piece anywhere on the board in place of his move, presuming the resulting state of the board does not leave the king in check. The most common piece used here is the "pocket knight."

 * "Pawn odds." The stronger player, usually as black (to compound the difficulty), removes a single pawn from his starting position, usually the pawn on f7 (the king's bishop pawn). In one set of three games played by Kasparov against an amateur, two pawns were removed. My dad has an interesting story about how his opponents in high school gave him eight pawn odds (removed all of them), accelerating their development, which he claims made him a stronger defender and tactician when the pawns came back on the board.

 * "Piece odds." In increasingly severe forms of handicap, the Queen's Knight, the Queen's Rook, or the Queen herself is removed from the board of the stronger player at the start of the game.

More Common Handicaps Compatible with Orthodox Chess

 Some handicaps are compatible with orthodox chess. These include:

 * First move. Since the late nineteenth century, this has meant taking the white pieces; historically, the first move could be given to white or black. This handicap is interesting because it is unavoidable, particularly in any matchup consisting of an odd number of games (1, 3, 5, etc.).

 * Time controls. The weaker player could be allotted more minutes on the clock, permitting him or her to spend more time thinking, or, in blitz or faster games, increasing dramatically the odds of winning on time.

* Draw odds. The draw in chess, which is usually scored as 0.5 for both players, could be scored as a win for the player getting the handicap. This doesn't seem to change the outcome of the game as much as the scoring of it; however, knowing that a draw is as good as a win may change strategy for both players.

The first handicap is the only one commonly seen in rated chess, and it is usually balanced out somehow to allow all players a fair amount of time with the white pieces. However, in a (somewhat controversial) form of tournament tiebreaker known as "Armageddon," all three handicaps are involved: black is given more time on the clock in a blitz game where a draw for black counts as a win.

A common handicap utilized by masters in unrated games is the simultaneous exhibition; the master will play a room full (5 to 50) of more novice players by travelling around the boards and playing each game in turn. Usually there is no clock involved, and the master, if he wishes, can pass on a turn if it becomes at all "interesting" in order to return to the game when most of the other exhibition matches have completed.

One of the most universally applicable observation about all handicaps is this: The severity of a handicap increases as the skill of the weaker player increases. If the weaker player is at 800, a player at 1200 might suppose to play without a queen. If the weaker player is at 1200, the stronger player should probably be at 1800 or higher; and if the weaker player is at 1800, he probably has a fair (or more than fair) chance to win against anyone (even a 2800 rated GM) who played without a queen on the board.

It is estimated that the first move handicap accounts for approximately 50 Elo points at moderate to high levels of play; in other words, a 1650 player with the black pieces is as good as a 1600 player with the white. Based on the above observation, one expects this handicap to account for more at the very highest levels of play, i.e. letting white 2400 be equal to black 2500. From the reverse direction, one expects that the handicap becomes almost negligible at lower levels of play, as seems to be confirmed by empirical evidence; players under 1000 don't seem to have their wins as strongly correlated to first move advantage as higher rated players.

Do We Need More Handicaps?

In light of the above, it would appear that chess has generated enough forms of handicap for the purposes of the friendly game between players of disparate skill. However, there is a problem: most of the more severe handicaps are incompatible with orthodox chess. With just first move and more time on the clock (and often with just first move because many chess programs can't assign different clocks), even the difference between a 1200 player and a 1500 player is stark. This is unfortunate because there are times when these types of players may wish to have a friendly game in which the outcome is in serious doubt.

There are other less common ways to introduce handicaps within the bounds of orthodox chess. Of the three that come to mind, two require a pledge from the stronger player because none of the rules of orthodox chess require the stronger player to play in such a way as to provide the handicap once the game has started.

The one that doesn't require such a pledge is the one that mucks with the opening. For example, on the second move or the third move the stronger player can be required to move his king. This would be more than a waste of tempo; it would prevent castling. Or, if the stronger player has the white pieces, the second move can be required to be a3, thus nearly giving black two moves to respond to white's one.

One that does require a pledge is one where the stronger player agrees to "sacrifice" a designated piece within a certain number of moves, such as ten. The rule of the sacrifice should be that the piece that captures the sacrificed piece is not under threat on the subsequent move at the time of the announced sacrifice. 

For players of significantly different strength, a practice of the 16th through 19th century was the pion coiffé or "capped pawn" where checkmate must be delivered with a predesignated pawn (typically not a d or e pawn). If this pawn were lost, the stronger player could not win the game. Since this significantly changes the winning condition of chess, making it narrower than orthodox chess, it is not clear that it is entirely compatible. (There are many conceivable positions from which traditionally the only options would be checkmate--with other pieces--or to concede material or give the opponent drawing chances, perhaps even the win.)

Last but not least...

The goal of the handicap is to make a friendly between significantly different strength players more fun. For this reason, it is suggested that the handicap itself tend to increase the fun of the game. Although some variants such as "Pocket Knight" are fairly entertaining, at least for the weaker player, none of the handicaps within orthodox chess really do that so far.

This is where I will introduce my own version of a severe chess handicap within orthodox chess. This handicap can also be applied to both players leading to a fun game between equal strength playing partners. The handicap can also be made less severe, allowing two players to use different severity handicaps (or none) depending on their relative strength and preference. This handicap is to limit artificially the selection of moves with the use of a random input, such as the throw of a pair of dice. The rules (and there is more than one possible set) for this handicap will be published in a subsequent blog post.