Loading field map
Checker play is the branch of backgammon concerned with moving the checkers around the board. It is distinguished from the game's other major decision area, cube play, which governs the use of the doubling cube. While cube play asks whether the stakes of a game should be raised, checker play asks how the checkers should be moved. Every turn presents a choice among legal moves, and checker play is the art and science of choosing among them.
The central question of checker play is deceptively simple: given a particular position and a particular roll of the dice, which move maximizes the player's chance of winning the game? The difficulty arises because the answer depends on a complex interplay of factors: the race (the distance each side must travel to bear off), the structure of each side's home board, the positions of blots (exposed single checkers), the state of the match score, and the position of the cube. A move that is correct in one context can be a serious error in another.
All checker play decisions can be understood as a negotiation between two broad objectives: racing and fighting. Racing is the simple business of moving checkers toward the home board and off the board as quickly as possible. Fighting involves the tactical struggle for control of the board: hitting opponent blots, making points (occupying a point with two or more checkers so it cannot be landed on), and building a blockade of made points that impedes the opponent's progress.
Most positions call for a blend of both objectives, but the correct blend varies enormously. In a pure race, where both sides have cleared their home boards and no contact is possible, the only question is speed. In a contact position, where checkers can still be hit, the player must weigh the immediate gain of advancing against the risk of leaving a blot that the opponent can hit, and the strategic value of building structure that will pay off later.
The tension between racing and fighting is not a school of thought but the fundamental strategic axis of the game. Different positions, and different phases of the game, demand different emphases. The opening, the middle game, the bear-off, and the endgame each have their own characteristic problems.
The opening is the most studied phase of checker play. Because the starting position is fixed and the first roll offers only fifteen possible combinations (with the two dice indistinguishable), the correct opening moves have been analyzed exhaustively. The standard opening moves are well established: for example, a roll of 3-1 is played by making the 5-point (the opponent's bar point) with the 8-point and 6-point checkers, while a roll of 6-1 is played by making the bar point. These moves are not arbitrary; they reflect the deep principle that making points in one's own home board and outer board is the most efficient use of a roll.
The opening establishes the strategic character of the game. A player who makes several home-board points early has a strong defensive position and threatens to hit any blot the opponent leaves. A player who advances checkers into the opponent's outer board is racing ahead but may leave blots vulnerable to attack. The opening is thus the first arena where the racing-fighting tension appears.
The middle game is the phase where most of the game's complexity resides. Here, the board is partially built, both sides have checkers in various stages of development, and the possibilities for hitting, blocking, and escaping are numerous. Middle-game checker play is the subject of the most intense analytical effort, both by human experts and by computer programs.
A central concept in middle-game play is the prime: a sequence of consecutive made points that blocks the opponent's checkers. A prime of six consecutive points is a full blockade, trapping any checker behind it. Building a prime is a slow, structural strategy that sacrifices immediate racing speed for long-term control. The opponent's response to a prime is often to attempt to escape a back checker (a checker on the opponent's home board) by rolling a high number, or to hit a blot in the prime to break it up.
Another key concept is timing. Timing refers to the relationship between the number of checkers each side has in play and the number of points each side needs to make. A player with good timing has enough checkers in the outfield to continue building and improving the board while the opponent is forced to move. A player with poor timing may be forced to break up a strong position because all checkers have advanced too far. The concept of timing is subtle: it is not simply about having more checkers, but about having the right number of checkers in the right places to execute a plan.
The tactical side of middle-game play involves the calculation of specific sequences. When a player leaves a blot, the opponent must decide whether to hit it. Hitting is often correct, but not always: hitting may bring the opponent's back checker into play, may leave the hitter's own blot exposed, or may disrupt the hitter's own timing. The decision to hit or not is one of the most frequent and most difficult in the game.
The bear-off is the phase where both sides are bringing their final checkers home and removing them from the board. In a pure bear-off, with no contact possible, the play is purely a race, and the correct moves are determined by simple arithmetic: the goal is to bear off as many checkers as possible while minimizing the chance of leaving a blot that could be hit if contact is still possible.
When contact is still possible in the bear-off, the play becomes more delicate. A player may choose to leave a blot in a position that is safe from immediate hitting but that preserves flexibility for future rolls. The concept of safe versus bold play is central here. Safe play avoids leaving any blot that can be hit; bold play accepts the risk of a hit in exchange for a better racing position or better future flexibility. The choice between safe and bold is governed by the match score, the cube position, and the relative strength of the two home boards.
The endgame, when one side has borne off most checkers and the other is still trying to escape, is a phase of precise calculation. The player who is ahead in the race must decide how much risk to take to finish; the player who is behind must decide whether to leave blots in the hope of hitting a shot. These decisions are often close, and small errors can be decisive.
Checker play cannot be fully understood in isolation from the doubling cube. The cube changes the value of the game, and therefore changes the risk-reward calculus of every checker play. A player who is ahead in the match may be willing to accept a slightly worse checker play to avoid the risk of a gammon (a double win that counts as two points). A player who is behind may take greater risks to create a chance of winning.
The interaction between checker play and cube play is bidirectional. A checker play that creates a strong position may justify a double; a cube decision may be influenced by the checker play options available on the next roll. Expert players think of the two as inseparable, even though they are analyzed separately for pedagogical purposes.
The modern understanding of checker play has been shaped decisively by computer analysis. Beginning in the late twentieth century, neural-network-based programs such as TD-Gammon and later GNU Backgammon and XG Mobile were able to evaluate positions with a level of accuracy far beyond that of any human. These programs did not simply play well; they provided a way to measure the error of any given move by comparing it to the program's own evaluation.
The result has been a transformation of checker play theory. Many moves that were once considered standard have been shown to be errors, and many moves that were once considered unorthodox have been shown to be correct. The concept of the blunder—a move that costs more than a certain threshold of equity—has become central to the modern vocabulary of the game. Players now study with computers, reviewing their own games to identify blunders and understand why the computer's move was better.
This analytical revolution has not replaced human judgment but has refined it. The computer's evaluations are based on millions of rollouts (simulated continuations of the game), and they provide a standard of correctness that is objective in a way that human intuition can never be. However, the computer's evaluations are not always transparent: it is often difficult to understand why a particular move is best, even when the computer says it is. The modern player must therefore combine computer analysis with human understanding, using the computer as a tool for discovery and verification.
Within the practice of checker play, several distinct approaches can be identified, though they are not mutually exclusive schools so much as complementary modes of analysis.
The positional approach emphasizes the long-term structure of the board. Its practitioners think in terms of primes, anchors (made points in the opponent's home board), and timing. The positional approach is associated with the classical literature of the game, and it remains essential for understanding why certain moves are good or bad. Its limitation is that it is qualitative: it can identify the strategic themes of a position but cannot always determine the best move with precision.
The tactical approach emphasizes the calculation of specific sequences. Its practitioners look for hitting opportunities, escape routes, and immediate threats. The tactical approach is essential in positions where the game is volatile and a single roll can change everything. Its limitation is that it can be overwhelmed by the sheer number of possibilities in complex positions.
The equity-based approach is the modern, computer-driven method. It evaluates moves by their expected value, measured in terms of the probability of winning, losing, and gammon outcomes. The equity-based approach is the most accurate, but it is also the least intuitive: it often recommends moves that seem strange to human players because the computer has seen deeper into the position.
These approaches are not rivals but tools. A strong player uses all three: the positional approach to form a plan, the tactical approach to calculate the immediate consequences, and the equity-based approach to verify the choice and to catch errors that human intuition would miss.
The present landscape of checker play is defined by the coexistence of a rich human tradition and a powerful computational tool. The classical concepts—racing, fighting, timing, primes, safe versus bold—remain the vocabulary of the game. They are not obsolete; they are the framework within which the computer's evaluations are understood and communicated. At the same time, the computer has revealed that the classical framework is incomplete. Many positions that were once thought to be clear are now known to be subtle, and many moves that were once dismissed are now known to be correct.
The result is a field that is both deeply traditional and rapidly evolving. The best players in the world are those who can combine the strategic understanding of the positional approach with the precision of the equity-based approach. The study of checker play is the study of a game that is simple in its rules but inexhaustible in its complexity, and the modern player stands on the shoulders of both the classical theorists and the computer programmers who together have mapped its terrain.