Loading field map
Match equity is the branch of backgammon theory concerned with the value of a player's position in a match, expressed as the probability of winning the entire match from that point. It is the quantitative foundation for many of the most important decisions in match play, especially doubling decisions, which in backgammon are governed by the "match score" rather than by the raw value of the current position on the board.
In a single game of backgammon, the objective is to win the game, and the "cube" (the doubling cube) is used to raise the stakes. The decision to double or accept a double is governed by the "equity" of the game: the expected value of the game in points, given the current position and the cube value. In a match, however, the objective is not to maximize points but to win the match. A player who is leading in the match may be willing to accept a double that is unfavorable in terms of game equity, because losing a single point is less damaging than losing the match. Conversely, a player who is trailing may be willing to take a double that is unfavorable in game equity, because the chance of a large swing is worth the risk.
Match equity is the quantitative expression of this idea. It assigns a number to every possible match score (e.g., "2-away, 3-away" meaning the leader needs 2 points and the trailer needs 3) that represents the probability that the player to move will win the match, assuming both players play optimally. This number is called the "match equity" of that score. The match equity of a score is the fundamental input to all doubling decisions in match play.
The modern theory of match equity began in the 1970s, when backgammon was experiencing a surge of popularity and the first serious attempts were made to analyze the game mathematically. The earliest match equity tables were constructed by hand, using a combination of game theory and empirical observation. The most famous of these is the "Kazaross" table, published in the 1980s, which was based on a large number of computer simulations of matches. The Kazaross table was the first to be widely accepted as accurate, and it remains the standard reference for match equity in the backgammon community.
The development of match equity was closely tied to the development of the "doubling cube" itself. The cube was introduced in the 1920s, but it was not until the 1970s that the theory of doubling in match play was formalized. The key insight was that the value of a double in a match is not the same as the value of a double in a game. In a match, the value of a double is determined by the "match equity" of the resulting score, not by the "game equity" of the position.
There are two main approaches to match equity: the "theoretical" approach and the "empirical" approach. The theoretical approach attempts to derive match equity from first principles, using game theory and the concept of "optimal play." The empirical approach uses simulation and statistical analysis to estimate match equity from the results of many games.
The theoretical approach is based on the idea that match equity is a function of the "match score" and the "skill" of the players. The simplest theoretical model is the "linear" model, which assumes that match equity is a linear function of the score difference. This model is not accurate, but it is a useful starting point. More sophisticated theoretical models, such as the "exponential" model, attempt to account for the fact that the value of a point is not constant throughout a match.
The empirical approach is based on the idea that match equity can be estimated by simulating many matches and recording the results. This approach is more accurate than the theoretical approach, but it requires a large number of simulations to be reliable. The Kazaross table is an example of an empirical approach.
The two approaches are not mutually exclusive. In fact, the most accurate match equity tables are produced by a combination of the two. The theoretical approach provides a framework for understanding the structure of match equity, while the empirical approach provides the data to fill in the details.
The most important use of match equity is in the doubling decision. In a match, the decision to double is governed by the "match equity" of the resulting score. The player who is considering doubling must compare the match equity of the current score with the match equity of the score that would result if the double is accepted. If the match equity of the resulting score is higher than the match equity of the current score, then the double is a good decision.
The same logic applies to the decision to accept a double. The player who is considering accepting a double must compare the match equity of the current score with the match equity of the score that would result if the double is accepted. If the match equity of the resulting score is higher than the match equity of the current score, then the double should be accepted.
The match equity of a score is also used to determine the "take point" and the "double point" in a match. The take point is the minimum game equity that a player must have to accept a double. The double point is the minimum game equity that a player must have to make a double. These points are different from the take point and double point in a single game, because they are based on match equity rather than game equity.
Match equity is a powerful tool, but it has its limits. The most important limit is that it assumes both players play optimally. In practice, players make mistakes, and the match equity of a score is not the same as the probability that a player will win the match if the player is not playing optimally. The match equity of a score is the probability that a player will win the match if both players play optimally.
Another limit is that match equity is a function of the skill of the players. The match equity of a score is not the same for two players of different skill levels. A player who is much stronger than the opponent may have a higher match equity than the match equity of the score would suggest, because the stronger player is more likely to win the match.
Finally, match equity is a function of the "gammon rate" of the players. The gammon rate is the probability that a player will win a game by a gammon (a win by more than one point). The match equity of a score is not the same for two players with different gammon rates. A player with a high gammon rate may have a higher match equity than the match equity of the score would suggest, because the player is more likely to win a game by a gammon.
The modern landscape of match equity is dominated by the use of computer programs. The most important of these is the "GNU Backgammon" program, which is a free, open-source program that includes a match equity table. The program is used by many players to analyze their matches and to improve their play.
The match equity table used by GNU Backgammon is based on the Kazarator table, but it has been updated to reflect the results of more recent simulations. The table is not perfect, but it is the best available.
The use of match equity has also been extended to the analysis of "match play" in general. The concept of match equity is now used to analyze the "opening" of a match, the "middle game" of a match, and the "endgame" of a match. The match equity of a score is also used to analyze the "doubling cube" in the context of the match.
The future of match equity is likely to be dominated by the use of computer programs. The computer programs are becoming more accurate, and they are also becoming more accessible. The match equity of a score is a powerful tool, and it is likely to become even more important in the future.
Match equity is the theory of the value of a score in a match of backgammon. It is the foundation of the doubling decision in match play, and it is a powerful tool for analyzing the game. The theory of match equity is based on the concept of the probability of winning the match, and it is used to make decisions about doubling and accepting doubles. The theory is not perfect, but it is the best tool that we have for understanding the game of match play.# Match Equity
Match equity is the branch of backgammon theory that quantifies the value of a match score. In a match played to a fixed number of points, the score alone—independent of the current position on the board—carries a measurable meaning: it tells each player their probability of winning the match, assuming both play optimally from that point forward. This probability is called the match equity of the score. Match equity is the conceptual bridge between the raw arithmetic of points and the strategic decisions, especially doubling, that make match play different from a series of independent games.
In a single game of backgammon, the doubling cube multiplies the stakes. A player who doubles offers the opponent a choice: accept the double and play for twice the current stake, or resign the game and concede one point. The decision to double or accept is governed by the game equity—the expected value of the position in points. In a match, however, the objective is not to maximize points but to win the match. A point gained or lost has different consequences depending on the score. A player leading 4-away to 2-away, for example, may accept a double that is unfavorable in game equity because losing two points still leaves the match winnable, while declining would effectively end the match. Conversely, a player trailing may accept a double that is unfavorable in game equity because the chance of a large swing is the only realistic path to victory.
Match equity makes these trade-offs precise. It assigns to every possible score a number between zero and one: the probability that the player to move will win the match, assuming optimal play by both sides. This number is the foundation for all doubling decisions in match play. Without it, a player has no principled way to decide whether a double is advantageous at a given score, or whether a take is correct.
The doubling cube entered backgammon in the early twentieth century, but the systematic analysis of doubling in match play came much later. In the 1970s, as backgammon experienced a surge of popularity and the first serious attempts at computer analysis began, players and theorists started to construct match equity tables. These tables list the match equity for every possible score in a match to a given length.
The earliest tables were constructed by hand, using a combination of game-theoretic reasoning and simulation. The most influential of these was the Kazaross table, developed in the 1980s by Kit Woolsey and Hal Heinrich, based on extensive computer simulation of matches. The Kazaross table was the first to be widely accepted as accurate, and it became the standard reference for match play analysis. It remains the basis for most modern match equity tables, though it has been refined with more recent data.
The development of match equity was closely tied to the development of the "match equity" concept itself. The key insight was that the value of a double in a match is not the same as the value of a double in a single game. In a match, the value of a double is determined by the match equity of the resulting score, not by the point value of the game. This insight transformed the doubling decision from a purely positional calculation into a calculation that depends on the match score.
Two broad approaches to match equity have coexisted since the field's beginning. The theoretical approach attempts to derive match equity from first principles, using game theory and assumptions about how players' winning probabilities change with the score. The empirical approach estimates match equity by simulating many matches and observing the results.
The simplest theoretical model is the linear model, which assumes that match equity is a linear function of the score difference. This model is easy to compute but inaccurate, because it fails to account for the fact that the value of a point is not constant throughout a match. A point that brings a player to the winning threshold is worth more than a point that merely increases a lead. More sophisticated theoretical models, such as the exponential model, attempt to capture this nonlinearity by assuming that the probability of winning a game is a function of the score difference in a particular way. These models are useful for understanding the structure of match equity, but they are not accurate enough for practical play.
The empirical approach is more accurate. It involves simulating a large number of matches, using a computer program that plays at a high level, and recording the results. The match equity for each score is then estimated as the proportion of matches won from that score. This approach requires a large number of simulations to be reliable, but it produces tables that are accurate enough for practical use. The Kazaross table was the first such table, and modern tables are based on the same method, with more simulations and better playing programs.
The two approaches are not rivals in the sense of competing for acceptance. The theoretical approach provides a framework for understanding why match equity behaves as it does, while the empirical approach provides the accurate numbers needed for play. The most accurate tables are produced by combining the two: the theoretical model provides a smooth curve that is fitted to the empirical data, reducing the noise in the estimates.
The primary use of match equity is in the doubling decision. In a match, the player who is considering a double must compare the match equity of the current score with the match equity of the score that would result if the double is accepted. If the match equity of the resulting score is higher, the double is correct. The same logic applies to the decision to accept a double: the player must compare the match equity of the current score with the match equity of the score that would result if the double is accepted.
This comparison can be expressed in terms of the "take point" and the "double point." The take point is the minimum game equity that a player must have to accept a double. The double point is the minimum game equity that a player must have to make a double. Both are derived from the match equity table. In a single game, the take point is fixed at 25% (because the player who accepts a double risks one point to win two, so the break-even point is 1/3, but the cube's future value lowers it). In a match, the take point varies with the score. At a score where the trailer is far behind, the take point may be much lower, because the trailer's chance of winning the match is already small, and the double offers a chance to catch up. At a score where the leader is far ahead, the take point may be higher, because the leader can afford to decline a double and still have a good chance of winning the match.
The match equity table also determines the "redouble" decision. When a player owns the cube, the value of the cube is not just the current stake but the option to redouble later. The match equity of the score after a redouble must be compared with the match equity of the score if the redouble is not made. This calculation is more complex, but it is still based on the same match equity table.
Match equity is a powerful tool, but it rests on assumptions that are not always met. The most important assumption is that both players play optimally. The match equity of a score is the probability that a player will win the match if both players play optimally. In practice, players make mistakes, and the actual probability of winning a match from a given score depends on the relative skill of the players. A stronger player has a higher probability of winning from any score than the match equity table suggests, because the stronger player is more likely to capitalize on the opponent's mistakes.
Match equity also depends on the "gammon rate" of the players. A gammon is a win by more than one point, and the gammon rate is the probability that a player wins a game by a gammon. The match equity of a score is not the same for two players with different gammon rates. A player with a high gammon rate has a higher match equity at scores where a gammon is particularly valuable, because the chance of a gammon is a way to win the match more quickly. The Kazaross table and its successors are based on the average gammon rate of the players in the simulations, and they are not perfectly accurate for players whose gammon rates differ significantly from the average.
Finally, match equity is a function of the match length. The match equity of a score in a match to 5 points is not the same as the match equity of the same score in a match to 11 points. The tables are constructed for each match length, and the differences are small but not negligible.
The modern practice of match equity is dominated by computer programs. The most widely used is GNU Backgammon, a free and open-source program that includes a match equity table and a doubling analysis. The program uses a table based on the Kazaross table, updated with more recent simulations. Players use these programs to analyze their matches, to check the correctness of their doubling decisions, and to learn the match equity table for the scores they are likely to encounter.
The match equity table is also used in the analysis of match strategy beyond doubling. The concept of "match equity" is used to evaluate the opening of a match, the middle game, and the endgame. The table is used to determine the "match equity" of a position, which is the probability of winning the match from that position, given the score. This is a more general concept than the match equity of a score, and it is the basis for the "match equity" analysis in modern backgammon programs.
The future of match equity is likely to be shaped by the continued improvement of computer programs. The programs are becoming more accurate, and they are also becoming more accessible. The match equity table is a tool that is now available to every player, and it has transformed the way the game is played. The theory of match equity is not a complete theory of backgammon, but it is a fundamental part of the game's strategic foundation.