BGonline.org Forums

GnuBg settings to mimic XGR++

Posted By: Daniel Murphy
Date: Tuesday, 31 July 2012, at 1:55 a.m.

In Response To: GnuBg settings to mimic XGR++ (Taper_Mike)

Mike:

1) You're using version 0.90-mingw 20120408, I'm using Version GNU Backgammon 0.90-mingw 20111003.

2) The "noise" for one player was a mistake and happened because I had first set the rollout ply to change to "advanced" (which includes noise) after 2 plies, and then corrected that to change to 1-ply after 2 plies but neglected to remove the noise. I don't know what cause "the first 5995833 0-ply moves" instead of whatever a normal move filter is.

3) I also have a question about plys. GnuBg has a setting for “Evaluation for later plys.” Under that heading, I put a checkmark for “Enable separate evaluations,” and then I selected “Change eval ply after:” 2. This seems to be the setting that Daniel used. I wonder, though, if it is the right one.

Seems right.

4) XGR++ makes its “First 2 decisions in 3-ply, then 2-ply checker and 3-ply cube.”

That (and the 7-ply and trials) was all I was trying to simulate with Gnubg, and so the equivalent of "First 2 decisions in 3-ply, then 2-ply checker and 3-ply cube" should be "First 2 decisions in 2-ply, then 1-ply checker and 2-ply cube."

5) Rather than try to duplicate my and your efforts, here's what my first attempt should have looked like:

The score (after 0 games) is: White 0, Me-1 8 (match to 11 points)

Move number 5: Me-1 to play 53

White	174
Me-1	72
Position ID: zB5wQHDbzUMQAA Match ID: UYluAQAAQAAE

• Me-1 moves 10/5*/2

Cube decision
2-ply cubeless equity		+1.327626	(Money: +1.76373 +1.327626)
	0.941003 0.843050 0.040684 - 0.058997 0.001996 0.000004
Cubeful equities:
1.	No double	+1.156093
2.	Double, take	+1.048485	-0.107608
3.	Double, pass	+1.000000	-0.156093
Proper cube action:		Too good to redouble, pass

Alert: doubtful move ( -0.005)

#		Ply	Move	Equity
	1	R	15/10 6/3	+1.288662
			0.946018 0.886139 0.041785 - 0.053982 0.000718 0.000003 +1.362141 +1.288662 0.001048 0.001645 0.001579 - 0.001048 0.000099 0.000002 0.002728 0.003965
			Truncated cubeful rollout (depth 7) with var.redn.
			360 games, Mersenne Twister dice gen. with seed 691526212 and quasi-random dice
			Play: world class 2-ply cubeful prune [world class]
			keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
			Skip pruning for 1-ply moves.
			Cube: 2-ply cubeful prune [world class]
			Different evaluations after 2 plies:
			Play: 1-ply cubeful
			keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
			Cube: 2-ply cubeful
•	2	R	10/5*/2	+1.283891 ( -0.004770)
			0.949014 0.894383 0.069816 - 0.050986 0.000619 0.000003 +1.372639 +1.283891 0.001033 0.001889 0.001800 - 0.001033 0.000108 0.000003 0.002814 0.003976
			Truncated cubeful rollout (depth 7) with var.redn.
			360 games, Mersenne Twister dice gen. with seed 691526212 and quasi-random dice
			Play: world class 2-ply cubeful prune [world class]
			keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
			Skip pruning for 1-ply moves.
			Cube: 2-ply cubeful prune [world class]
			Different evaluations after 2 plies:
			Play: 1-ply cubeful
			keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
			Cube: 2-ply cubeful
	3	R	15/7	+1.277992 ( -0.010670)
			0.942554 0.888239 0.043668 - 0.057446 0.001967 0.000025 +1.354746 +1.277992 0.001137 0.001662 0.001803 - 0.001137 0.000367 0.000007 0.002943 0.004306
			Truncated cubeful rollout (depth 7) with var.redn.
			361 games, Mersenne Twister dice gen. with seed 691526212 and quasi-random dice
			Play: world class 2-ply cubeful prune [world class]
			keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
			Skip pruning for 1-ply moves.
			Cube: 2-ply cubeful prune [world class]
			Different evaluations after 2 plies:
			Play: 1-ply cubeful
			keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
			Cube: 2-ply cubeful

Messages In This Thread

• Stick's Matches from Novi
  Dmitriy Obukhov -- Monday, 30 July 2012, at 3:22 a.m.
  • Stick's Matches from Novi
    Stick -- Monday, 30 July 2012, at 4:07 a.m.
    • Stick's Matches from Novi
      David Wells -- Monday, 30 July 2012, at 8:19 a.m.
  • Stick vs Merriman: XG++ Data Makes No Sense
    Paul Weaver -- Monday, 30 July 2012, at 8:57 a.m.
    • Stick vs Merriman: XG++ Data Makes No Sense
      Carlo Melzi -- Monday, 30 July 2012, at 10:08 a.m.
      • Stick vs Merriman: XG++ Data Makes No Sense
        Bob Koca -- Monday, 30 July 2012, at 2:27 p.m.
    • Stick vs Merriman: XG++ Data Makes No Sense
      Rick Janowski -- Monday, 30 July 2012, at 3:09 p.m.
      • Stick vs Merriman: XG++ Data Makes No Sense
        Bob Koca -- Monday, 30 July 2012, at 4:04 p.m.
        • Stick vs Merriman: XG++ Data Makes No Sense
          Keene -- Monday, 30 July 2012, at 4:32 p.m.
          • Stick vs Merriman: XG++ Data Makes No Sense
            Bob koca -- Monday, 30 July 2012, at 4:53 p.m.
        • Stick vs Merriman: XG++ Data Makes No Sense
          Rick Janowski -- Monday, 30 July 2012, at 5:53 p.m.
    • RO and XG is having issues with this score and cube position it seems
      neilkaz -- Monday, 30 July 2012, at 3:42 p.m.
    • Stick vs Merriman: XG++ Data Makes No Sense
      Stick -- Monday, 30 July 2012, at 3:47 p.m.
      • Stick vs Merriman: XG++ Data Makes No Sense
        Rick Janowski -- Monday, 30 July 2012, at 5:48 p.m.
    • Stick vs Merriman: XG++ Data Makes No Sense
      Daniel Murphy -- Monday, 30 July 2012, at 6:27 p.m.
      • problem at truncation -
        eXtreme Gammon -- Monday, 30 July 2012, at 7:34 p.m.
        • problem at truncation -
          Daniel Murphy -- Tuesday, 31 July 2012, at 4:45 a.m.
      • GnuBg settings to mimic XGR++
        Taper_Mike -- Tuesday, 31 July 2012, at 12:36 a.m.
        • GnuBg settings to mimic XGR++
          Daniel Murphy -- Tuesday, 31 July 2012, at 1:55 a.m.
          • GnuBg settings to mimic XGR++
            Taper_Mike -- Tuesday, 31 July 2012, at 5:28 a.m.
            • GnuBg settings to mimic XGR++
              Daniel Murphy -- Tuesday, 31 July 2012, at 6:25 a.m.
              • GnuBg settings to mimic XGR++
                Taper_Mike -- Tuesday, 31 July 2012, at 10:58 a.m.