[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [computer-go] future KGS Computer Go Tournaments - two sections?

On 10, May 2005, at 10:42 PM, Don Dailey wrote:


Hi David,

I think it's my fault in the way I worded things that maybe sounded
like an attack on SlugGo.   I apologize if that is how it came across.

No, I do not feel that I or the program has been attacked. Neither
by you or anybody else in this discussion. It is important for us to
figure out what the fair and proper procedure should be.


The key phrase in my statement about not wanting to play several GNU
Go variants was "IN THE SAME TOURNAMENT."

So the way to understand my stance is that I don't want to play
several GNU Go clones in the same tournament, but I would be happy to
play any single one.

Thanks for the clarification.


And yes, I would be very pleased to be part of your verification
process, but botnoid is probably too weak for you to mess with.  In my
opinion you need to play a long public match against GNU Go, MFOG or
some other strong program on KGS or one of the other servers.

We do this in our own lab. I am wondering what the additional benefit
is for doing it on KGS. Games against GNU Go are particularly unfair.
SlugGo's lookahead scheme is too good at predicting what GNU Go
will do (we call it evil twin syndrome) and thus we crush GNU Go far
out of proportion to our actual difference in strength. For other
opponents the lookahead is not as effective because of a low hit rate
on their moves.


Tournaments are pretty much useless (or at best would require
significantly more effort to get enough game samples to actually
impress a scientist.)
Yes, I agree. The statistics of hundreds of Go games is very interesting,
and when the programs we have compared get close in strength, either
because they are close to begin with or by adjusting the handicap, the
variance gets very large.

The only conclusion I can reach after having seen the results of several
thousand games between these programs is that one tournament is far
too short to be a meaningful representation of relative playing strength
when the programs are "close." I am still attempting to quantify "close."



Cheers,
David


_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://www.computer-go.org/mailman/listinfo/computer-go/