Firstly, thanks for such a great program.
I'm sure that the folks at gnubg headquarters run tests where gnubg,
Snowy
and Jellyfish are left overnight to fight it out. So I'm sure you
know by
now what, for example, is the expected amount of money that gnubg
makes or
loses when playing a single money game for $1 a point against Snowy,
when
they are both allowed to average 1 second per move on the same
machine, or
something like that. Why not publish these statistics? I note that
Snowy and
Jellyfish don't enter the Computer Games Olympiads, so I'm guessing
that
they run these simulations, and they know that they would probably
lose. Or
maybe it's very close, and they think that they would lose more
sales by
losing than they would gain by winning, since their measure is real
money.
Adam.