First Cube of the 2009 OSBC Open Final Match

(c) 2009 by Mary Hickey

When I began studying my final match of the 2009 OSBC Open event, I was surprised that Snowie 4’s 3-ply evaluation said Neil Garvie had a take when I doubled here. It said the money equity was .574 and his pass was wrong by .021. (In people-speak, those numbers mean I had a mandatory double and he had a close pass.)

Over the board, I had been thinking after my crushing 33 the previous roll he had a drop, but maybe the 44 he had just rolled would be enough to get him to incorrectly take. It wasn't and he didn't! But here was Snowie saying he should have. It was rollout time, for sure.

I first tried a 1-ply "Mini Rollout", which expanded the money equity for my side to .597. This made it a very skinny take, by .010, for money but a pass at the score by .037. The 3-ply evaluation found a similar difference between money and the actual score of 0-0/11, saying the take for money was right by .066 compared to .021 in the match. I then performed the rollout you see here, which gave a result similar to the Mini, showing that Neil was correct to pass this cube.

 

Why the difference between money play and the beginning of a moderate-length match? Because not all points are of equal value in a match--the later ones are the biggest ones! The distortions become greater toward the end of a match, but they exist at the low end also.

Here, the risk of losing four points by getting yourself gammoned is more immediate than your chance of turning the game around and winning four points on a recube. At this score, that block of four points is worth slightly more than four times the value of the single point you can pass now. For money, they're all the same. That difference is small, but it is real!

Now, on to what is going on with the position itself. We all know an ace point game by itself isn't a take for money or at normal match scores, but that if there are some extra ways you can win, it might make the cut. This position isn't quite good enough to be an "ace point with extras", partly because of insufficient extras, but also because the ace point game Neil may end up playing may not be well-timed.

Neil is forming his board in good order without wastage, but that isn't an "extra". Rather, it is expected, or the equity of the ace point game would be eroded, making the need for extras even greater. The 21-anchor game he might form would be better than what he has, but still not enough for a take because I can easily attack and frequently succeed in stopping him from making it.

What would give Neil a take here? If he had either his five or bar point already made, that would be enough of an improvement to give him an easy take both for money and at the score.

What we can learn from this problem:

1. When Snowie says a cube is a take on 3-ply, but it looks shaky to your human eyes, then run, don't walk, it to the Rollout Court of Appeals.

2. A takeable "ace point with extras" game needs to start with at least a reasonably timed ace point game to fall back on if the extras fail. If you don't have that, you need a good chance to carry out a different game plan. If that also isn't available to you, the game may be too gammonish for you to take.

3. The score of 0-0/11 isn't "just like money", though it is close. With that said, we know that 0-0/9 and 0-0/7 can't be "just like money" either. A precise explanation of all the differences is beyond the scope of this article (and also beyond the scope of what I know about the game, for that matter). But for starters, at an equal score early in a match, try being a little more circumspect than for money about taking gammonish cubes.

If you are a member of GammonVillage, you can view the entire match at this link.
You can also view a report on the tournament itself here.
 

Many thanx to Mary Hickey for this article - published 16th April 2009