BGonline.org Forums

64R-64, data

Posted By: Nack Ballard
Date: Tuesday, 19 October 2010, at 8:16 p.m.

In Response To: 64R-64 4 ply RO (neilkaz)

64R-64

Congratulations, Neil! AFAIK, you're the first to place P on top in a long 64R-64 rollout. Here is data (first four rows are money, last two are AtS):

[S P9 Z56] ~20*5 (ext in progress)
[S P7] /46
[S=P] <62
[P S3] ^46

d[S P16] /5
g[P S18] /5

Key: S = Split (24/18 13/9), P = Point (8/2 6/2), Z = reverse split (24/20 13/7). Error sizes are in thousandths.

Before the brackets, d = dmp, g = gammon go; else assume money. After the brackets, the bot symbols are: ~ Snowie, / Gnu, < XG 3-ply, ^ XG 4-ply. This is followed by the number of trials in thousands, rounded down (and an asterisk is a repeater).

For example, "d[S P16] /5" means that at dmp, S is best and P is -.016 in a GnuBG rollout of 5000+ trials.

The XG 4-ply rollout result notwithstanding, I prefer playing S at all scores that are not GG-ish, except against opponents that I have experienced as having special trouble playing properly against the 2pt.

Nack

Messages In This Thread

• 64R-64 4 ply RO
  neilkaz -- Tuesday, 19 October 2010, at 11:57 a.m.
  • 64R-64, data
    Nack Ballard -- Tuesday, 19 October 2010, at 8:16 p.m.
    • 64R-64, data
      neilkaz -- Thursday, 21 October 2010, at 1:47 a.m.