Play to Win

What play would you make ?

play-to-win-1
Good backgammon players have the ability to calmly take risks at appropriate moments. Or in other words, they would rather play to win rather than play to avoid losing badly.

This position illustrates this idea. Black is in a poor position. Had he rolled a 5 he would have had reasonable chances of winning but instead the dice have given him a 62 to play.

No matter what he does he will be leaving shots and at least two blots for white to attack. He could make a 5-pt point board with plays like 13/7, 6/4 or 15/9, 6/4 but both of these plays leave multiple shots and that nice home board will soon disintegrate as soon as several black men end up behind white’s broken five-point prime.

The ‘play to avoid losing badly’ approach is to play 15/9, 15/13. This will leave only two blots but any 4 by white still more or less wins the game and white will often attack the blot on his 2-pt when he is unable hit the outfield blot. Even if black survives this roll he will still be in trouble and the initiative lies with white.

Can that initiative be taken over by black? Yes it can. It may seem suicide but the correct play in this position is actually 13/5*! Can a play that exposes four blots possibly be correct?

Once again the answer is yes and for the following reasons:
Black has a four-point home board, just as strong as white’s, so he should seek to utilise that asset.

The main reason is that on 16 (out of a possible 36) rolls next turn white will languish on the bar If black survives that next roll he will have very real winning chances and will even win a gammon on occasion. As an example look how 33 for white plays after 13/5* and 15/9, 15/13 – quite a marked difference!

This is the ‘play to win’ approach. Black has been given the chance to seize victory from the jaws of defeat and should take that opportunity.

play-to-win-2
As you can see from the rollouts below the decision is not close. 13/5* is clearly correct. It loses about the same number of gammons as the apparently ‘safer’ 15/13, 15/9 but wins an extra 10% of the games – which is a very big difference.