The rules of backgammon are on their own page, and the doubling cube and match play are where they belong. What follows is something else: how to think about a position once the rules are known.
One decision that comes back every move
Every move in backgammon is a decision made on incomplete information: the dice have not been rolled yet. And the single most common strategic decision in the game is always the same one — intensify the conflict between the checkers, or reduce contact and turn the game into a race.
That is why a game plan is not chosen once. It is chosen again from the position in front of you, and the task is closer to picking from a set of known options than to inventing something.
Three ways to win
There are three routes to a win — racing, attacking and priming — and every other plan reads as a combination of them.
- The running game. The plan of the player who is ahead: bring the checkers home and bear them off with as little interaction with the opponent as possible. It has no positional complexity, and that is exactly what separates it from the rest.
- The attack, or blitz. Aggressive assault: hit the opponent’s checkers repeatedly, keep them on the bar and close your home board so there is nowhere to come back to.
- The prime. An unbroken wall of consecutive points with the opponent’s checkers trapped behind it.
- The holding game. You keep an anchor in the opponent’s home board — a point there, or occasionally his bar point — to hinder his bear-in and to hit a blot he leaves while clearing his outfield points.
- The back game. Two or more points in the opponent’s home board and a bet on a late hit. It gets its own section below: it is the most disputed plan in the game.
The pip count: the number a plan is tested against
The most important single number in backgammon is the pip count, and the detailed count exists to answer one question — who is ahead in the race. If your count is lower, you are ahead.
It is worth counting for a practical reason. There are two uses, and the second matters more than the first: the pip count guides your checker play and, more importantly, your cube decisions. Knowing your count against your opponent’s tells you directly whether to double, take or drop in a pure race — and in a pure race, with no contact, cube decisions are the most straightforward ones there are.
Keep counting to the end: even while bearing off, the pip count tells you whether you are ahead or behind. A rough estimate works too — if most of your checkers are already halfway round while your opponent still has many back, you are ahead.
Racing: when to run and when to slow down
The advice repeated most often, and in the most languages, fits into four words: when ahead in the race, race. Avoid contact and take your checkers home.
The other half of the same rule is less obvious: when you are behind, slow down. Racing while you are losing the race means losing it to the end, so the player who is behind looks for contact instead of avoiding it.
How the race ends is a subject of its own: bringing the checkers home and taking them off without spending the advantage is set out on the bear-off page.
The five-point and the opening
Points are not worth the same: the five-point is worth more than any other single point on the board, and it matters most in the early game. Not for one reason but for three at once — it anchors your own home board, it limits your opponent’s re-entry from the bar, and it starts a potential prime. Next in value come the four-point and the bar point, the seventh; making the bar point begins a prime and cramps the opponent’s back checkers.
On the board the value of those points is immediate.
The three most valuable points on the board are marked. Held in a row, they are already a wall — which is why the bar point is called the start of a prime.
Opening rolls set the structure of the whole game, which is why the best plays for them are simply worth memorising. After that there is nothing left to memorise: once the first move is made, positions become unique, and the plan takes over from memory.
Particular rolls worth knowing by heart:
- 3-1 is the best opening roll. It makes the five-point (8→5 and 6→5) and leaves you holding three points at once: the six, the five and the eight. Making the five-point on the first move is called the single best opening play in the game.
- 6-1 makes the bar point: a checker from the thirteenth point and one from the eighth both go to the seventh. By strength, though, it is only the third best roll — behind 3-1 and 4-2.
- 6-3 is played as a split: one back checker goes from the twenty-fourth point to the eighteenth, and one from the thirteenth to the tenth. Running a single checker to the fifteen-point gives up almost nothing in strength, but it is the harder and more dangerous play.
One more opening rule: hitting your opponent’s blots in the opening is almost always right. And a separate technique — the quickest way to make a point is to slot it first, accepting the risk of being hit.
The anchor: where to stand in the opponent’s home board
An anchor is a point you own in your opponent’s home board, and the best square for it is settled: the twentieth point — the opponent’s five-point: it is the hardest point for him to attack and it forces him to play very carefully. That anchor is called golden.
What separates the golden anchor from any other is where it sits, and where it sits is only visible on the board.
The golden anchor: White holds a point inside the opponent’s home board, and it is the hardest point there to attack. While the anchor stands, no complete prime can be built against White.
An anchor also defends against a prime: while you hold one, a complete prime cannot be built against you. An anchor on the opponent’s one, two or three point cuts his gammon threat significantly — at a cost in position.
Deciding when to break the anchor is among the most important decisions in the game, and the mistake is expensive: breaking it and running after your opponent has rolled well for two turns loses the race and the safety net together.
Splitting your back checkers pays off: it improves your chances of getting an advanced anchor and your coverage of the opponent’s outer board.
Priming and blitzing: how checkers get trapped
A prime is an unbroken wall of consecutive points with the opponent’s checkers shut behind it. A blocking structure works better the more checkers are stuck behind it — against two or more it works, against one it barely does. The same holds for a prime: it is most effective when your opponent has two or more checkers trapped. A full prime — six points in a row — cannot be crossed at all: no roll of the dice jumps that wall.
The word “wall” is not a metaphor here, and the board shows it literally.
A full prime: six points held in a row (marked), with two black checkers shut behind them — nothing jumps the wall. Against a single trapped checker the same wall would cost nearly as much and give almost nothing.
A blitz works differently: repeated hits, an attempt to keep the opponent on the bar, and a closed home board so that re-entry is impossible. Hitting a checker and sending it to the bar gains you something precisely when your home board is already strong — otherwise it comes straight back. A successful blitz puts the opponent under fresh pressure: an early double forces them to either accept at a bad score or err while short on time to build their board.
Caution stays part of the tactics, and short-term safety ranks high. Two blots in your own home board is already too many: your opponent can hit both and take a massive advantage. Blots in his home board without a reason are avoided too.
Duplication and diversification
Two techniques with unfamiliar names, both about the numbers on the dice rather than about the checkers.
Duplication means placing your checkers so that your opponent needs the same numbers for different jobs. His effective choice narrows, though he rolls just as often.
Diversification is the mirror image applied to yourself: spread your checkers out to increase the number of good rolls on your next turn.
The back game: a last resort and what it costs
A back game is a plan where you build a strong position in your own home board while leaving blots in your opponent’s; in another wording, you secure at least two points in his home board and hope he is forced to leave a shot late in the game. Moving into one means building points in his home board.
Here is what it looks like while the plan has yet to pay off:
A back game: White holds two points in the opponent’s home board (marked) and waits for him to be forced to leave a shot. The cost of the plan is visible in the same picture — those checkers are the farthest from home, and if the shot never comes the loss is a heavy one.
There is no settled view of the back game. Some players call playing one from the start almost always ineffective and keep it as a last resort for a player who is far behind. Others hold that a well-timed back game turns a catastrophe into a win in positions that looked lost.
What they agree on is this: the hardest part of a back game is the timing. And on the warnings:
- do not attack if you have a lot of work to do in your own back game;
- two deep points in your opponent’s board are not a back game yet, and labelling it one too early is a mistake — at the early stages two points are not enough;
- a back game that fails often ends in a gammon.
The cube: where counting ends and judgement starts
The mechanics of the cube are on its own page; here only the decisions.
The cube is called one of the most powerful strategic weapons in the game, and the general rule is short: it is usually right to double when you have the advantage. The signs of that advantage, the ones on which a double is offered: a prime built, more of your checkers home, the opponent on the bar.
The reference numbers:
- take a double when you have a realistic chance of winning — about 25 % or more;
- drop when your opponent’s chance of winning is above 75 %, that is, better than three to one;
- when both players are two points away from winning, take if you expect to win more than 32 % of the time — roughly one game in three.
There is also a rule about doubt, and it is neatly built: if you are unsure whether to double, double — because giving your opponent the decision gives him the chance to get it wrong. The strict form of the same principle is known as These numbers have a far side, the one seen by the player offering the cube. It is called the doubling window, and its edges are blurred: doubling is advised at winning chances of roughly 60–75%, the window is sometimes widened to 50–80%, and 66–75% is often named as the band where doubling is right and taking is right too. The spread shows the window is an area of judgement, not a line at which a rule switches on. Its early edge carries a mistake of its own: doubling too soon is what gives the opponent the chance to take and run.
Woolsey’s rule: if you are not sure whether your opponent has a correct take, you must double.
The cube belongs in holding games too, where the gammon chances are significant. And it works as pressure: an opponent can be trapped into taking a cube in a bad position, or bluffed into dropping so that you cash the game quickly.
The Jacoby rule changes the behaviour of a player with a large lead: it pushes him to double and end the game rather than play it out hoping for a gammon. The game speeds up as a result — the positions where nobody doubles because they are playing for a gammon simply stop happening.
The match score changes the correct play
This is probably the least obvious part for anyone who has only played single games. In a match the correct play often depends on the score, and tactics adjust to the score rather than to the board.
- The value of a gammon at a given score is different for the two sides — one player needs it, the other has almost no use for it.
- After the Crawford game the trailer should double immediately at the start of every game.
- If your opponent is one point away from winning the match, you will want to double as early as possible in order to catch up.
- The Crawford rule itself changes strategy radically once it comes into force.
Separately: playing to save a gammon is often correct — a lost game is not always a finished one.
What players argue about
Besides the argument about the back game there is a second one, and it is about temperament.
Some hold that the most aggressive player is the favourite in backgammon. Others put short-term safety distinctly higher. A third view reconciles the two: experts take a lot of risk, but the genuinely strong player always weighs the gain against the level of risk.
The argument is not about a fact but about where to set the baseline.
How to learn this
Three pieces of advice to close with.
Play against the computer: it is the best way to practise. Memorise the best opening plays, so you do not solve again what is already solved. And a habit worth forming early — look for the best play of the six available, even though it will not always be the best one, because if you do not go looking for it you may not spot it.
And in Narde?
Narde strategy lives on its own page. The game there is a different one — no hitting, one checker off the head — and almost nothing said here carries over: the attack, the anchor and the back game are simply not possible.
Comments