Louisville Backgammon Club

Frank’s Position Analysis — March 2007

by Frank Frigo

Backgammon can be a game of pattern recognition. We have all seen opening positions so many times in both live games and when employing our “bot” friends that it is easy to assume that the game doesn’t get very complicated until we are well into the middle game. Not only can early positions become very complicated, but they can greatly influence the outcome of a game and match. It is best to take your time when playing your early rolls and avoid falling into the trap of habitual plays.

On a recent Tuesday night, I was playing Quint McTyeire in the second round of our club tournament, and things got interesting very quickly. After Quint opened by making his 5-point with a 3-1, I responded with a 5-1 (position 1). Most players split in this position as I did. Splitting 24/23 and playing 13/8 puts pressure on Quint’s stripped 8-point while diversifying my back checkers in hopes of making an advanced anchor or increasing my prospects for a future hit in the outfield.

Quint replied with a 6-5 and made the standard running play with 24/13 (position 2). 24/18, 13/8 also deserves merit here and may well be the best play by a slim margin. It invites action when Quint has the stronger board and reinforces the stripped 8-point. After I made my 4-point with a 4-2, Quint had another interesting play with 6-2 (position 3). I must admit that I would have run my remaining back checker 24/16 without much thought while trying to take advantage of an 11 pip lead. Quint, however, chose to play 13/5 and he may well have made the best play. He brings another important builder into the zone of his home board while I still have some consolidation to consider. He also keeps his rear checker out of harm’s way for the moment.

Now things began to get really interesting. I followed with a 2-1 (position 4), and after considering the elegant anchor play of 24/22, 23/22, I talked myself into playing 24/23, 13/11. My reasoning went something like this:

1. I secure an anchor, maybe not the best anchor available, but an anchor nonetheless.
2. He only has one man back, so I would like to get another builder working toward my 5-point or bar.
3. The 23-point anchor may immobilize his stripped 8-point, and besides, I need to button up now that he has that extra builder on his 5-point.

A rollout indicates that my play is a clear error. Quint’s 8-point becomes a very effective blocking point when I anchor on the 23-point as opposed to the 22-point. He is happy to leave the checkers on his 8-point just where they are and focus on making the bar – in fact, with his next roll of 4-3 (position 5) that is exactly what he correctly did.

My next roll of 2-1 (position 6) provides the most complicated problem of this young match. Just eight rolls into the first game, we have already faced several interesting choices. This decision is particularly difficult because of my error on the previous turn. Quint’s position has become very strong as he has several useful checkers in range of his 4-point and bar point. I decided that hitting on the ace point was necessary to take away half of what would likely be a constructive roll for Quint. With the 1, I decided to split to where I should have been previously – on the 22-point. This may seem like a very wide open play, but I have to do something strong here. I cannot simply let him prime me out of the game. Rollouts concur with my play although many alternatives are reasonable. Here is a ranking of the first five choices after a 3-ply rollout of 360 trials (variance reduction, precise speed) with Snowie 4:

1. 23/22, 6/1*
2. 11/10, 6/1 (-.06% MWC)
3. 13/8, 6/5 (-.14% MWC)
4. 13/7 (-.17% MWC)
5. 8/7,6/1* (-.32% MWC)

It is easy to see from this game just how quickly things can take shape early in a match and how much our seemingly subtle choices can affect future decisions.

– Frank Frigo

Return to Features