by Douglas Zare
1 September 2005

If you simply play well, you will win a lot. Bots do this. To have the
most success, you should be willing to make theoretical
mistakes in order to provoke larger errors from your opponents,
particularly when you have a significant skill advantage.
When I played on GameSite 2000 (now
GammonSite) from 2000-2001, I usually had the highest rating of any human player on the server by over 100 rating points. This is explained primarily by the lack of competition, but my rating sometimes exceeded that of the Snowie bots. I wasn't nearly as strong as Snowie at that time, but I was better than Snowie at exploiting the mistakes of the weaker human players, particularly in matches of 3-7 points. Part of my advantage came from studying how aggressive or cautious to be with the doubling cube at each match score against weaker opponents
| The rest of this article (11.76 K) is premium content. |
Article text Copyright © 1999-2010 Douglas Zare and GammonVillage Inc.
Douglas, what about the reverse? What if I am an intermediate player against a world class player? In a recent tournament I (intermediate) was paired against a world class player in a side event. I was down 1-3 in a 5 point match. I handled the cube as if it were a post Crawford game - cubing immediately. I now see that this was a mistake as nothing would have lost my market that soon. But how should I have handled the cube? Bob F.
In general, weaker players should apply the modifications in the opposite direction.
Although it was a technical mistake to double immediately at 2-away 4-away, it is not nearly as large of a mistake as you might expect. Against an equal opponent, doubling immediately converts the score to Crawford 2-away, and the trailer wins only about 1% less at Crawford 2-away. Against a stronger opponent, many players would be better off doubling immediately at 2-away 4-away, since they tend to make large mistakes by forgetting to double later. Of course, you might do better still to consider the cube carefully at each opportunity, but your double was not a large error.
Douglas Zare
Douglas this article reminds me a match we played a few month ago on FIBS. It was a 5 point match, and you started in the first game with an early 55 putting me two checkers on the bar, after my play, you doubled and it was a perfectly take (confirmed by snowie) but I losed a gammon (between 25-30% risk of losing a gammon when you doubled). Knowing that you are better than me with a huge difference ( rating 1800 versus 1950) should I had not take this gammon risk and dropped your double ?
Sorry, I don't remember that match. Nevertheless, if you feel you are a weaker player, you should be happy to take a more aggressively at 5-away 5-away.
Consider the entries in table C for the scores we might reach. A skill advantage applies most to 4-away 5-away and 5-away 4-away, and least to Crawford game. There isn't enough of an opportunity for a skill advantage to change the outcome of a very lopsided match. If you are a weaker player, it is more valuable to reach these scores than it is normally.
Another thing to consider is the position. It can be hard to blitz correctly, while it is easy to dance correctly. (On the other hand, most blitzes fail, and then it may be harder to counterattack properly than to play the failed blitz.) Players who are normally weaker may have a skill advantage when getting blitzed, though this depends on the position and players. That also argues for taking.
It's important to remember to redouble aggressively at 5- away 5-away. Not only is the take point higher than normal, but there is much more skill left in the match at 3-away 5-away than in the Crawford game.
Douglas Zare
You must be signed in to post comments.

