Backgammon Cube Handling

Cube Handling

Liveliness of the cube

From:   Kit Woolsey
Address:   kwoolsey@netcom.com
Date:   7 April 1997
Subject:   Re: Liveliness of the cube
Forum:   rec.games.backgammon
Google:   kwoolseyE88w6E.B1C@netcom.com

> I have a question concerning the liveliness of the cube. In a money > game, assuming a dead cube the takepoint is 25%; assuming a completely > alive cube the takepoint becomes 20%. Common wisdom says that the real > takepoint is something like 21.5% (corresponds to an equity of -0.57). > This would mean that one considers the cube to be alive to 70%. How > can we compute mathematically this number? No, we can't. It depends on the type of positions. Two examples: a) A long race. Here the trailer is likely to have a reasonably efficient recube if things go his way, so the cube is relatively alive. b) A holding game -- way behind in the race (so must hit a shot), but the trailer having a perfect prime which he will be able to maintain for several rolls. In this case, if the trailer does hit his shot he will suddenly become a huge favorite and lose his market by a mile. Thus, the cube is relatively dead. The above are examples of positions where gammons are impossible or very unlikely. If gammons are in the air, things get more complex still. For example, consider a blitz position. Assuming the trailer has a close pass/take decision, this means that his win percentage is way above 25% (to compensate for the many gammons he loses). Thus, he will be able to get in even more recubes than normal. Also, this type of position tends to lead to very efficient recubes when the game turns around, since usually the trailer makes his improvements slowly. Therefore, the cube is even more alive than normal. The 21.5% figure is the generally agreed figure for the "average" position. However, the actual number definitely depends on the position in question. > Also, in some scores for example 3-away 4-away the live takepoint is 19% > but the dead takepoint is around 34%, what is considered to be the real > takepoint? Instinctively, I would guess that the recube of the trailer is > more alive than usual due to the very high takepoint of 40% of the > redouble. Again, I would be very interested in how one could compute this > number. Yes, the cube is more alive than usual due to the takepoint of the redouble. And, once again, for the reasons above, there is no way to compute this number. Kit

Did you find the information in this article useful?

Do you have any comments you'd like to add?