Tuesday, June 19, 2007

Are good players just luckier?

Even at a substantial skill disadvantage, one should still win a lot of the time. One nice thing about backgammon is that you can make errors and bounce back; this also means that after making many errors, you will still be penalized for making more. In games without chance you should have
already lost.

Strong players still make a lot of mistakes. They are just smaller and less numerous than the errors of weak players. Even Snowie makes mistakes, but it's hard to be sure she is wrong.

Some errors in checker play arise when the rules of thumb conflict. (Should you hit, or make another inner board point? Should you make a bold play while ahead in the race but with a stronger board? Should you make a solid prime, escape, or start a blitz?)

Some errors arise when the rules of thumb make no suggestions. Some positions are so complicated or subtle or strange that the heuristics do not help, and stronger players may find it worthwhile to steer towards these knowing that the intermediate players will make more errors. The
point of the right play may be to gain timing rather than a racing advantage or any fixed asset. Often, there are several good plays, and one must choose based on intuition.

Some errors arise when the rules of thumb are wrong. Sometimes "ugly" plays are right, and stronger players are better at judging when this is the case. Sometimes the "pretty" plays have no winning game plan; they will just look good for a bit before collapsing.

In addition, intermediate players are often terrible with the cube. Tip for intermediates: play JF a lot, rapidly, or watch the demo mode. Get a feeling for when JF doubles, and double at least at those times against a human player.

Backgammon is surprisingly deep. The classic arsenal of heuristics can be learned in a week, so it is predictably insufficient.

Whether I qualify as a strong player depends on your definition. My rating on FIBS jumped up recently after I read some of Robertie's books, but every recent move from my matches that I entered into Snowie was pronounced a blunder. (One double, which my opponent dropped, required a 64% chance for my opponent to drop for it to be correct. I guess doubling at -2:-4 is
dangerous even with a 6-prime and my opponent's ace-point made, with very low gammon chances.) On the other hand, maybe that means that I now know which positions I'm confused about, rather than being so confident of erroneous plays that I don't check them.

No comments: