| |
BGonline.org Forums
Opening AtS data
Posted By: Nack Ballard In Response To: XG Rollout of Opening 64 at Gammon Save (Paul Weaver)
Date: Friday, 30 July 2010, at 10:04 p.m.
The original GnuBG rollout (that you never saw) of opening 64 s[S R4.8] 31k was corrupt. The extended portion, which by deduction was s[R S6.3] 15k, can be considered legitimate; these pieces average to the (two-thirds corrupt) rollout of s[S R1.1] 46k currently posted at this site.
Other rollouts done for this position on GnuBG:
64 s[R S6.5] 62k
64 s[R S7] 31k supremo
64 s[R S3.7] 5k supremoI know of no XG rollout (on either ply) that has yet been done for 64 at GS.
Thanks to David Rockwell for his XG 3-ply rollout results of the other positions you requested. His list is shown below. I have added three longer rollouts performed by Neil (see clickable links).
41 g[$ S10] 5k, and g[$ S9] 23k, and 4-ply g[$ S2] 46k
51 d[$ S1 R14] 5k
51 g[$ S11] 5k
62 g[$ S15 D27 R34] 5k
63 d[S R5 U29 Z30] 5k
63 s[R=S] 5k, and 4-ply s[S R6] 12kIf you extend any of David's early game files, please post results here at bgonline.
For reference/comparison, here is GnuBG data for the same positions (combined -- some on site, some not), rounded to nearest thousandth:
41 g[$ S8] 38k
51 d[S=$ R10] 46k 15kR
51 g[$ S10] 20k
62 g[$ S10] 25k (excl. prob-corrupt 46k on site)
63 d[S R6.2] 15k (possibly corrupt)
63 s[S R2] 5k supremo (excl. prob-corrupt 46k on site)Key: R = Run, S = Split, Z = reverse split, $ = Slot, D = Down, U = Up. Error sizes are in thousandths. Before brackets, d = dmp, s = save, g = go. Number of trials is in thousands (k), rounded down. For example, "g[$ S15 D27 R34] 5k" means that at (gammon) go, $ is best, S is -.015, D = -.027 and R is -.034 after 5000+ trials.
Nack
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.