| |
BGonline.org Forums
Updated data for 41-s
Posted By: Nack Ballard In Response To: 41-s ^46 RO (Nack Ballard)
Date: Thursday, 10 February 2011, at 4:39 p.m.
Thanks for the "<77 rollout, Neil. Here is data for opening 41 at save:
s[S $9 U27 T45] /15*9 5
s[S $21] ^46
s[S $10 U25] "<77*5Key: S = Split (24/23 13/9), $ = Slot (13/9 6/5), U = Up (24/23 24/20), T = sTack/Tower (13/8). Before brackets, s = save. After brackets is a bot symbol ( / GnuBG, ^ XG1 4-ply, "< = XG2 3-ply) and the number of trials in thousands, rounded down (an asterisk repeats a number).
For example, "s[S $10 U25] "<77*5" means that at (gammon) save, S is best and $ is -.010 in an XG2 3-ply rollout of 77000+ trials, and U is -.025 after 5000+ trials.
For reference, the GnuBG data of s[S $9 U27] /15*9 is a combination of two rollouts: s[S $12.3 U26] /10*3, and s[S $2.1 U27 T45] /5 (supremo).
Well, I guess that answers that question -- whether XG2 3-ply is more aligned with XG1 4-ply (not) or with GnuBG (though data for the latter totals only 15k for now).
Neil, in this post, referring to the XG2 3-ply rollout of 77k trials that finished at [S $10], you said, "41-s started for a while, expecting S to beat $ by over .020." Do you remember about how many trials had been rolled out at that point?
Nack
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.