BGonline.org Forums

NY Doubles Finals Cube #2 Rollout

Posted By: Matt Cohn-Geier
Date: Monday, 16 February 2009, at 7:21 p.m.

In Response To: NY Doubles Finals Cube #2 (Matt Cohn-Geier)

OTB when this came up I was very unsure what the correct cube action was...Take? Drop? Too good? Not good enough? My first inclination when I have no idea is to cube and let them figure it out but that has problems at scores like this where the cube comes back almost immediately and we get into trouble, or we cash a position that's far into TG land.

Let's see what's going on if we cube. If they drop, they're at -2, -7, easy enough. If they take they will rewhip on virtually any reasonable sequence, although once in awhile they will never get there and we'll gammon them at 2 for the match. Let's just assume they rewhip immediately. Then our gammons don't count and we're basically playing GS, although very few gammons against us in this position. Their gammonless takepoint in that scenario is about 31%. They can do quite a bit better than immediately rewhipping, though, since they should only cube when they have a threat of a market loss, GNU gives their takepoint at around 23% after accounting for gammons (we reach TG at 80%).

OTOH we can play on and try to get to Crawford or -2 -7 the other way. Stick keeps trying to convince me that this is a practical play-on even though the RO shows it to be a clear take. Arguments in favor of play-on: 1) I'm not very accurate at assessing these situations (no more so than anyone else) and could easily have gotten it wrong in any direction, in which case doubling with a window this narrow costs us more than waiting, 2) I don't really give that many people enough credit to take something like this. Arguments against playing on: 1) it's a take now, and if they're passing I'd be happy to just add another % or 2 to my MWC and run with the points, 2) with the doubling window so narrow, it's easy to lose our double if virtually anything goes right for them. Of course at that point we can just play the game as if it were cubeless but winning matches consists of raising your MWC in these situations.

So at the end of it all, I'm still entirely unconvinced as to what is technically best (clear take, but is it a cube?) and to what is practically best (no idea here). I had a longer RO (5kish) which was about 1 JSD in favor of ND/T, but I lost that. SW ROs will follow when they finish.

The score (after 0 games) is: Ray & Petko 0, MCG & CJC 4 (match to 7 points)

Move number 2: MCG & CJC on roll, cube decision?

Ray & Petko	150
MCG & CJC	160
Position ID: qHPMA0BsO2AAMw Match ID: cAngAAAAIAAA

• MCG & CJC doubles

Cube decision
Rollout cubeless equity		+0.8098	(Money: +0.9886)
Cubeful equities:
1.	Double, take	+0.8755
2.	Double, pass	+1.0000	+0.1245
3.	No double	+0.8479	-0.0276
Proper cube action:		Double, take
Rollout details
Win W g W bg Lose L g L bg Cubeless Cubeful Centered 1-cube 0.7241 0.5213 0.0733 - 0.2759 0.0513 0.0029 +0.8098 +0.8479 Standard error 0.0030 0.0043 0.0047 - 0.0030 0.0016 0.0005 0.0110 0.0209 Player Ray & Petko owns 2-cube 0.7322 0.5665 0.1161 - 0.2678 0.0523 0.0051 +1.2116 +0.8755 Standard error 0.0047 0.0071 0.0075 - 0.0047 0.0016 0.0013 0.0258 0.0396
Full cubeful rollout with var.redn.
885 games, Mersenne Twister dice gen. with seed 817744508 and quasi-random dice
Play: 2-ply cubeful prune [world class]
keep the first 0 0-ply moves and up to 12 more moves within equity 0.16
Skip pruning for 1-ply moves.
Cube: 2-ply cubeful prune [world class]

Cube decision
Rollout cubeless MWC		83.193%	(Money: +0.9886)
Cubeful equities:
1.	Double, take	83.620%
2.	Double, pass	84.430%	0.810%
3.	No double	83.441%	-0.179%
Proper cube action:		Double, take
Rollout details
Win W g W bg Lose L g L bg Cubeless Cubeful Centered 1-cube 0.7241 0.5213 0.0733 - 0.2759 0.0513 0.0029 83.193% 83.441% Standard error 0.0030 0.0043 0.0047 - 0.0030 0.0016 0.0005 0.071% 0.136% Player Ray & Petko owns 2-cube 0.7322 0.5665 0.1161 - 0.2678 0.0523 0.0051 85.806% 83.620% Standard error 0.0047 0.0071 0.0075 - 0.0047 0.0016 0.0013 0.168% 0.257%
Full cubeful rollout with var.redn.
885 games, Mersenne Twister dice gen. with seed 817744508 and quasi-random dice
Play: 2-ply cubeful prune [world class]
keep the first 0 0-ply moves and up to 12 more moves within equity 0.16
Skip pruning for 1-ply moves.
Cube: 2-ply cubeful prune [world class]

Messages In This Thread

• NY Doubles Finals Cube #2
  Matt Cohn-Geier -- Sunday, 15 February 2009, at 5:40 a.m.
  • NY Doubles Finals Cube #2
    Christian Munk-Christensen -- Sunday, 15 February 2009, at 9:32 a.m.
    • NY Doubles Finals Cube #2
      Stick -- Sunday, 15 February 2009, at 9:41 a.m.
      • NY Doubles Finals Cube #2
        Christian Munk-Christensen -- Sunday, 15 February 2009, at 5:40 p.m.
        • NY Doubles Finals Cube #2
          Bill Riles -- Sunday, 15 February 2009, at 5:42 p.m.
          • NY Doubles Finals Cube #2
            garyo -- Monday, 16 February 2009, at 6:25 a.m.
  • NY Doubles Finals Cube #2
    Martin -- Sunday, 15 February 2009, at 10:14 a.m.
  • NY Doubles Finals Cube #2
    Bill Riles -- Sunday, 15 February 2009, at 2:31 p.m.
  • NY Doubles Finals Cube #2
    neilkaz -- Sunday, 15 February 2009, at 4:40 p.m.
  • NY Doubles Finals Cube #2
    garyo -- Monday, 16 February 2009, at 6:23 a.m.
  • NY Doubles Finals Cube #2 Rollout
    Matt Cohn-Geier -- Monday, 16 February 2009, at 7:21 p.m.