Here’s a pair of very similar positions that shows how an (apparently) small change in an early-game position can dramatically change your approach.

Cash game. White to play 5-1

Cash game. White to Play 5-1

These two positions show a routine situation in the early middle game. Both sides have an anchor, and the race is very close. White’s position is slightly preferable since  he has better chances of establishing a block against Black’s anchor.

White’s roll of 5-1 is a pretty routine shot and he has two obvious choices: 9/3, starting the 3-point, and 13/8 9/8, playing safe while keeping all checkers is front of Black. Of these two plays, 9/3 is slightly better; 13/8 9/8 creates a big stack which will prove awkward in some of the upcoming variations. In the middle game, we try to avoid creating big stacks unless alternatives are really weak. There’s nothing structurally wrong with 9/3, so it’s a better play here.

There is, however, yet another alternative: the bold and dashing 13/7, slotting the bar point and leaving two blots, but forcing Black to give up his anchor if he wants to hit. Is 13/7 just a crazy play, or a stroke of genius?

If you found yourself a spectator at backgammon tournaments in the 1970s and early 1980s, you would have seen plays like 13/7 occur with some frequency. Slotting to build primes was the essence of the “pure style” much in vogue at the time. Advocates of this style would have made several compelling arguments in favor of 13/7:

> If Black doesn’t hit, White can build a prime quickly.

 > If Black does hit, he loses his anchor and White can attack. The attack might still let White build a formidable position.

 > If White gets several men sent back, he can construct a back game/blocking game formation. Black’s board isn’t very strong, and White already owns a high anchor, so Black has little or no chance of forming a prime against whatever formation White can construct.

 > The positions that will result from 13/7 will all be more complex than those that will arise after the more standard plays. The better player will have an edge in these positions.

 > Most of the good players play this way, and they’re winning all the big tournaments, so don’t you want to play like them?

Actually, these are all pretty good arguments. While they’re not good enough to tell you that the play is right, they are good enough to tell you that the play might be right.

As the 1980s wore on and turned into the 1990s, experience accumulated and plays like 13/7 became less popular, even among the good aggressive players. A general sense took hold that these plays were a little too over-the-top, although they remained a useful weapon against weaker players.

The arrival of the bots and their rollout capacity in the mid-1990s gave us a tool that let us put plays like 13/7 under a microscope. The results were interesting and can be summed up pretty quickly:

> In normal positions, where Black has reasonable structure, double-slotting plays like 13/7 are mostly wrong. (Although this is backgammon, and as always there are exceptions.) Just how wrong they are depends on the strength of Black’s position. Where Black’s home board is not too imposing, as in Position (a), the slotting plays are only slightly wrong. Against a sufficiently weak opponent, the slotting plays might indeed be optimal.

> If Black’s game is very weak, as in Position (b), where Black has a blot in his 1-point board, a big stack on the 6-point, and a stripped 8-point, then the aggressive slot is likely to be correct. In many cases Black will have to avoid hitting even if he throws a three or a five, making the slot a no-brainer.

Solution:

 (a) 9/3.

 (b) 13/7.