Re: [Bug-gnubg] Measuring performance levels

[Morten Wang]

* Joseph Heled
> at 0ply, 12960 games I got
>
>   0.07% 3.6% 77.06% 23% 7.08% 0.0%
>
> So this is the same as the above (SN 3?).

I get about the same numbers, with variance reduction:

Rollout:
Win     W(g)    W(bg)   L(g)    L(bg)   Cubeless        Cubeful
0,7717  0,0409  0,0008  0,0705  0,0001  +0,5145         n/a
0,0005  0,0004  0,0005  0,0004  0,0055   0,0013         n/a

(still 21600 games, non-truncated, rotate rolls, seed 11, 0ply,
 cubeless, no Jacoby).


Morten!

-- 
"...a liquid which was almost, but not quite, entirely unlike tea."
  My opinions are not necessarily those of my provider,
  not necessarily mine, and probably not necessary.