Back to "The Zare Formula" written by Douglas Zare

Using All The Points

by Douglas Zare
22 December 2001

Douglas Zare

When a match is far from over, checker play and cube actions should be close to the analogous decisions for money, except for the Jacoby Rule. When one or both sides is close to the end, or when the cube escalates, match play can be quite different from money play. Where is the transition?

One heuristic, often cited in match analyses and used at times by Jellyfish is that match play is like money play as long as both players can use all of the points. For example, at 5-0 in a match to 11, 6-away 11-away, the idea is that the leader's initial double is as in money play since the leader can use the 2 points of a doubled win, and the leader can double based on gammon threats because the leader can use all 4 points of a gammon. The trailer may consider that the recube potential is normal because each side could use the full 4 points of a recube. However, if the leader is considering recubing to 8 at 6-away 11-away, this is not as powerful of a threat, since the leader can only use 6 points. If the leader recubes to 4, gammons are not as valuable, so they are not as threatening. Is this heuristic correct?

In this article we will see that this heuristic is sometimes valid, and sometimes quite wrong. The result is that some match play analyses based on this heuristic are wrong, and so is the way that Jellyfish plays at certain match scores. On the other hand, much of the time in long matches the heuristic is valid, so you don't need to know many entries of a match equity table. You can simply remember certain important scores, and when the heuristic applies, and how to handle the exceptions

The rest of this article (91.00 K) is premium content.