Before reading take at least a couple of minutes to decide how you would play the role and equally as importantly, why.

Now let’s look at the relative merits of each play:

(a) gets one checker out from behind white’s mini-prime (note that 9 of black’s numbers don’t let him move any of the checkers on his 21-pt). It  also creates a five-point prime against white’s lone back checker. On the downside any 6 by white will be a winning roll as he holds the doubling cube and can probably play on for a gammon with little risk. White does have some horror numbers in 55, 54 and 44 and 22 isn’t wonderful.

(b) also creates a five-point prime but has the added advantage of putting white on the bar against a five-point board. On the downside of this play is the fact that there are still three black checkers partially trapped on white’s 4-pt.

(c)    solves the problem of the three trapped checkers by releasing one of them. It gives white some additional bad numbers. Now 64 and 61 also leave black a direct shot. The downside of this play is that if white rolls a 2 he will have a winning position unless black re-enters immediately.

So which is the correct play? As with many decisions in backgammon it is a question of balancing risk against reward. You need to balance your experience with some detailed analysis such as we have done above to reach the right conclusion.

I got this wrong over the board by playing (b) which is actually bad enough to be a blunder! I was deceived by the opportunity to put a checker on the bar.
There are two key points about (a) and (c):

•    They both escape a back checker. Computers have taught us the importance of escaping from behind even small primes such as this one.

•    They allow white to leave another blot exposed with quite a few numbers. This in turn leads to black winning more gammons - a factor that the majority of people will have overlooked when doing their analysis.

(c) is much better than (a) because:

It is much easier for black to make a full prime after this roll than (a); it wins more gammons and it loses fewer gammons. In fact it is the clear winner by a huge margin – just look at the rollout data. I’ll wager not more than one in five readers will have got this problem correct.

Chris Bray
03/2007.