On Tue, 2008-10-28 at 11:40 -0400, Michael Williams wrote: > If you look at my original email, I did not mention removing that > until after I showed and discussed the numbers. And even then, I said > "... it is probably > safe to remove...". Probably being the operative word indicating that > I had not yet tested that.
I used your pseudo code, which didn't have the check in it. In tests I had done in the past, it didn't really make much difference. I had to run a lot of game to conclude that having the test seemed to make it slightly stronger and I don't think I ran enough to make this very conclusive. In general, I stop testing when it becomes clear that something won't make a big enough difference to worry too much about - then I use some other criteria - my gut feeling, simplicity, aesthetics, etc. But that is probably not good to do. Imagine ignoring a dozen 10 ELO enhancements because you were not willing to get enough data to prove each was a tiny improvement. You may have thrown away 100 ELO over time. - Don
signature.asc
Description: This is a digitally signed message part
_______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
