Remembering, Interpreting, to Learn from Past Errors
Paul Magriel, 1979
New York Times, March 15, 1979
To determine the correct play in a given position, a player will rely on past experience in similar situations. Accumulating such experience is an essential part of the learning process. All too often, however, players misinterpret events and learn the wrong lesson.

The diagrammed position, taken from a chouette at the Park 65 Backgammon Club in New York, illustrates how easily the wrong idea can be “learned.”

Black to play 3-3.
Black (the team captain) had just redoubled White and now had 3-3’s to play, Playing 8/2, covering the 2-point with two of the 3’s was easy. The team had conflicting ideas about the remaining two 3’s.

One player felt the best winning method was to attempt an immediate closeout by completing Black’s home board as soon as possible. With this in mind, he counseled playing 13/10, 8/5, bringing a builder into the home board in order to attack White on the 1-point as soon as White reenters.

A second player suggested 18/12, in order to get maximum outfield coverage. Rather than attempt to close White out immediately, he wished to make sure that Black would be able to contain White, should White reenter and run off the 1-point.

(a) 8/2, 13/10, 8/5
(b) 8/2, 18/12
The team captain rejected both these game plans. Over the vociferous objections of both his teammates, he took a third course, playing 13/7, putting a single man on the bar-point (7-point). He hoped to be able to make this point and so end further resistance by forming a full 6-point prime.

Unfortunately, White next rolled the dreaded 1-6, which enabled him both to reenter and hit the man Black had just left open on the bar-point. White won easily — in fact, Black was fortunate to avoid losing a double game.

In the post mortem, both his teammates were critical of his decision to “pay off” to this roll of 1-6. “Why take any chances when the game is all but won?” they complained. Black reluctantly agreed that his play had proved costly, and vowed to play more conservatively in the future.

(c) 8/2, 13/7

Black had, in fact, made the correct play, 8/2, 13/7. Sometimes it is best to take a small immediate risk (2 chances in 36) to insure victory rather than delay and face a potentially greater risk later on. In a roll or two, White will probably reenter on the 1-point. It is essential for Black to try to make the 7-point before White reenters. After White comes in it will be much more dangerous.

In order to make the point “naturally,” that is, without slotting, Black will probably have to wait many rolls. In this time, White will have had several opportunities to leap out with 6-6 or else hit one of Black’s outfield men with an indirect shot. The cumulative chances this gives White to win far outweigh the risks of an immediate 1-6. Paradoxically, because White’s home board is strong, this play is all the more necessary, because Black will always be afraid to attack White later, inside his home board.

The ability both to remember and correctly interpret past experience is not easy. White’s roll of 1-6 may be remembered for a long time because of its tremendous shock value. Because Black recalls this one roll so vividly, he is apt to give it disproportionate weight when considering slotting the bar in a similar position. On the other hand, if Black plays “safe” and then later gets into trouble and loses the game, he is unlikely to associate it with his earlier failure to slot.

XG logo
Tom Keith 2013 
Money play
White owns 2-cube
Black rolls 3-3

1296 games with VR
Checker play: 2-ply
Cube play: 3-ply Red

3-3: Game BG   Equity
1 18/12, 8/2 W
+0.6913 x  (b)
2 13/7, 8/2 W
+0.6458 (0.0455)  (c)
9 13/10, 8/5, 8/2 W
+0.4813 (0.2100)  (a)

Previous Column
March 8, 1979
Next Column
March 22, 1979

Main page for Magriel's NYT Columns

Index to the Columns

More articles by Paul Magriel

Backgammon Galore : Articles