Hi,
I did a simple experiment in which I let gnubg play against itself 1800 games in $ mode (unlimited) and without cube (actually, I played the first 20 games). This is an important point. The machine was set as computer against human, in settings>players and I had to click "end game" every single game. I noticed that the advantage of gnubg steadily increased. Intrigued, I did some statistical analysis and found a statistically significant difference in the probability of winning a gammon between gnubg and the human player. Apparently gnubg has a probability nearly 3% higher. I have attached a PDF file with my results. I find difficult to believe that gnubg cheats but I wanted to share with you these results, perhaps
someone knows better. Thanks!
FS