Once a week or so, I google Quackle to see what people have said about
it or done with it, and this is probably the most interesting thing
I've come across so far:

http://reason.cs.uiuc.edu/eyal/papers/scrabble-ijcai07.pdf

Mark Richards and Eyal Amir of UIUC took the source code from an early
release of Quackle and created two new players. One is essentially a
cheater that has and uses full knowledge of the opponent's rack, and
the other makes inferences as to what its opponent is likely to have
held. The Full Knowledge Player beat our now-retired Strong Player in
59% of its games, and the Inference Player won 51.4% against Strong
Player. I started working on inferences again shortly before the
Toronto Open, and I hope that we can do a little better than this
mainly by initially considering more plays, but their result is very
encouraging.

John O'Laughlin

Reply via email to