Re: [Bug-gnubg] new contact net

[Joseph Heled]

Albert Silver wrote:
> That's fantastic news! I announced it in GammonLine already. I presume
> you may still be testing it, but once finalized, (if it isn't already
> that is) I was wondering whether it would be possible to run tests to
> compare the overall improvement of the new nets over the old. Not just
> the contact net, but prior to the newer race and crashed nets. One
> question: which MET was used for testing, and what were the parameters
> tested? I presume these were at 0-ply considering the quantity, but
> asking is better than presuming. :-)

Yes, 0ply throughout. Not only this makes it possible to get the 
necessary large number of trials, but it is the best measure of the raw 
 quality of the net.

I used Snowie 2.1 MET, but from earlier checks it makes no difference at 
all.

-Joseph