Skip to first unread message


Nov 3, 2005, 6:47:45 PM11/3/05
to playhex
I finally installed Debian Linux on my machine, so one of the first
downloads was Six. It plays about the same level as Hexy, as far as I
can tell. The strongest level averaged more than 2 minutes per move on
a 14x14 board. The next level down averages about 30 seconds per move
on my machine, which is only 500 MHz, so you may do better. It suffers
from the same problem Hexy has- it refuses to play the whole board.
Wherever I would play, Six would almost invariably respond in the same
region. But I do like the greater range of grid sizes (up to 15x15),
and other options such as choosing how swap is dealt with. Its opening
book is not so good; It always plays 1.B1, even on the smaller grids
where it should be strong enough to see how bad that move is. It's not
just that 1.B1 has been proven to lose, it loses BADLY. When I go
first, it never swaps my 1.A3, which may be the correct decision. I
haven't experimented with other moves yet. It did beat me once on an
8x8 grid.

Anyway, I can now lay to rest any paranoid delusions I may have had
that some LG players were using Six to beat me. (There's no "infinite
time" setting.) ...


Nov 6, 2005, 2:13:02 PM11/6/05
to playhex
I'm not sure I agree with the characterization about refusing to play
the whole board,
but it is very much like Hexy.

I never noticed it playing B1 all the time because I don't use it that
way. I checked
it out though, and it seems you're right. How odd. I don't think it
has an opening
book beyond that, and in fact that first move may not come from a book.
It's pretty
clear all subsequent moves are worked out algorithmically, so if it's a
book it has only
a single entry.

I've been playing c3 a lot on Little Golem, and I like the results
because most of the
time it gets swapped and I win anyway. But I like my chances on both
sides. If I'm
right, then a3 is even more a "don't swap" opening.

++ kevin

Reply all
Reply to author
0 new messages