BGonline.org Forums

Something at stake -- rollout

Posted By: Daniel Murphy
Date: Wednesday, 11 June 2008, at 4:26 p.m.

In Response To: Something at stake (Caryl Stephanson)

I wonder why neither player has thrown the cube here. [...] Wasn't there somewhere it was shoulda coulda woulda? Yes, I know it's DMP.

When I say "double match point" I mean any score at which the current game decides the match with the current cube value. One-away all, 0-0/2 with a 2-cube, 5-3/7 with a 4 cube and 8-3/11 with an 8 cube are all double match point scores, and cube actions (and gammons and backgammons) are irrelevant.

Caryl must be thinking of DMP as "2-away all." It's true we sometimes call that score DMP but that usage either assumes that the game will become DMP with a double and take either immediately or before market loss, or must redundantly label a position "DMP" while simultaneously showing a 0-0/2 score and a centered cube.

Anyway, this is "DMP."

Double Match Point. Blue to play 5-3.

White	161
Blue	103
Position ID: tNZCgBnMnt0AAA Match ID: cIkuAAAAAAAA

• Blue weighed clearing the 8 point later against two anchors versus clearing it now and creating a two point gap, and chose 9/1 because it "looked right."

Alert: doubtful move ( +0.000)

#		Ply	Move	Equity
•	1	R	9/1	+0.316
			0.658 0.328 0.055 - 0.342 0.060 0.003 +0.316 +0.316 0.005 0.018 0.012 - 0.005 0.015 0.002 0.010 0.010
			Full cubeful rollout with var.redn.
			180 games, Mersenne Twister dice gen. with seed 828720450 and quasi-random dice
			Play: supremo 2-ply cubeful prune [world class]
			keep the first 0 0-ply moves and up to 16 more moves within equity 0.32
			Skip pruning for 1-ply moves.
			Cube: 2-ply cubeful prune [world class]
	2	R	8/5 8/3	+0.270 ( -0.046)
			0.635 0.339 0.061 - 0.365 0.058 0.008 +0.270 +0.270 0.005 0.021 0.012 - 0.005 0.010 0.004 0.011 0.011
			Full cubeful rollout with var.redn.
			180 games, Mersenne Twister dice gen. with seed 828720450 and quasi-random dice
			Play: supremo 2-ply cubeful prune [world class]
			keep the first 0 0-ply moves and up to 16 more moves within equity 0.32
			Skip pruning for 1-ply moves.
			Cube: 2-ply cubeful prune [world class]
	3	R	10/5 9/6	+0.257 ( -0.059)
			0.629 0.389 0.108 - 0.371 0.058 0.002 +0.257 +0.257 0.006 0.019 0.016 - 0.006 0.013 0.001 0.013 0.013
			Full cubeful rollout with var.redn.
			180 games, Mersenne Twister dice gen. with seed 828720450 and quasi-random dice
			Play: supremo 2-ply cubeful prune [world class]
			keep the first 0 0-ply moves and up to 16 more moves within equity 0.32
			Skip pruning for 1-ply moves.
			Cube: 2-ply cubeful prune [world class]
	4	R	10/5 6/3	+0.245 ( -0.071)
			0.622 0.365 0.090 - 0.378 0.046 0.006 +0.245 +0.245 0.005 0.018 0.013 - 0.005 0.007 0.004 0.011 0.011
			Full cubeful rollout with var.redn.
			180 games, Mersenne Twister dice gen. with seed 828720450 and quasi-random dice
			Play: supremo 2-ply cubeful prune [world class]
			keep the first 0 0-ply moves and up to 16 more moves within equity 0.32
			Skip pruning for 1-ply moves.
			Cube: 2-ply cubeful prune [world class]
	5	R	10/7 10/5	+0.237 ( -0.079)
			0.618 0.359 0.089 - 0.382 0.075 0.006 +0.237 +0.237 0.006 0.019 0.013 - 0.006 0.014 0.007 0.012 0.012
			Full cubeful rollout with var.redn.
			180 games, Mersenne Twister dice gen. with seed 828720450 and quasi-random dice
			Play: supremo 2-ply cubeful prune [world class]
			keep the first 0 0-ply moves and up to 16 more moves within equity 0.32
			Skip pruning for 1-ply moves.
			Cube: 2-ply cubeful prune [world class]

Evaluations. Gnubg clears the 8 point at 0- 1- 2- 3- and 4-ply.

Rank	9/1	8/5 8/3	10/5 9/6	10/5 6/3	10/7 10/5
RO	1	2	3	4	5
0-Ply	5 (-0.071)	1	3	2 (-0.031)	4
1-Ply	5 (-0.044)	1	2 (-0.015)	4	3
2-Ply	2 (-0.031)	1	5	3	4
3-Ply	2 (-0.007)	1	4	3	5
4-Ply	2 (-0.002)	1	5	3	4

If the score here were 2-away all with a centered cube, Blue should double and White should take. But if Blue doesn't double White should leave the cube in the middle. He has no market losers and should be happy to let Blue play on with a centered cube if Blue wants to do that.

At 2-away all with a centered cube Gnubg plays 8/5 8/3 on 0- 1- 2- 3- and 4-ply but in a rollout Gnubg will, I think, say that 8/5 8/3 is an EMG blunder at that score.

Messages In This Thread

• Something at stake
  Daniel Murphy -- Tuesday, 10 June 2008, at 1:16 a.m.
  • Something at stake
    Dave Ranson -- Tuesday, 10 June 2008, at 8:26 a.m.
  • Something at stake
    motiv4u -- Tuesday, 10 June 2008, at 9:58 a.m.
  • Something at stake
    Achim -- Tuesday, 10 June 2008, at 12:51 p.m.
  • Something at stake
    lucky player -- Tuesday, 10 June 2008, at 8:30 p.m.
  • Something at stake
    Caryl Stephanson -- Tuesday, 10 June 2008, at 8:48 p.m.
  • Something at stake
    Keene -- Tuesday, 10 June 2008, at 9:04 p.m.
    • Something at stake
      Caryl Stephanson -- Wednesday, 11 June 2008, at 2:07 p.m.
      • Something at stake -- rollout
        Daniel Murphy -- Wednesday, 11 June 2008, at 4:26 p.m.
        • Learning all the time as Benny Hill would say.
          Caryl Stephanson -- Wednesday, 11 June 2008, at 6:27 p.m.
          • found my theme song
            Caryl Stephanson -- Thursday, 12 June 2008, at 12:46 a.m.
      • Something at stake
        Keene -- Wednesday, 11 June 2008, at 4:31 p.m.
        • Something at stake
          Matt Cohn-Geier -- Thursday, 12 June 2008, at 5:00 a.m.
          • that it is NT *NM*
            Caryl Stephanson -- Thursday, 12 June 2008, at 3:18 p.m.