BGonline.org Forums

2-away 4-away cube decision

Posted By: ah_clem
Date: Monday, 16 April 2012, at 2:13 p.m.

In Response To: 2-away 4-away cube decision (ah_clem)

As expected, nobody had any trouble with the cube decision here. AtS the double is very clear even though as Neil points out it's not even a double for money. It's the take decision that's interesting.

I was pretty certain of the take OTB, thinking that passing when I still had very good winning chances would just be leaving points on the table. gnu 2-ply analysis backs me up on this indicating that passing is a whopper:

 

Cube analysis 

2-ply cubeless equity  +0.258 (Money:  +0.330) 

  0.617 0.173 0.007 - 0.383 0.081 0.002 

Cubeful equities: 

1. Double, take         +0.901 

2. Double, pass         +1.000  ( +0.099) 

3. No double            +0.729  ( -0.171) 

Proper cube action: Double, take 

In a post match discussion with my opponent, his gnu 3-ply analysis said it was a pass. Most of us are aware of gnu 3-ply's cube problems, so I put it down to that. I was quite surprised when a full 1296 2-ply rollout said pass, but only by .005 which was within the margin of error.

The only thing to do at this point was to finally break down and send some cash to Xavier for XG. And even XG has trouble with this one. 4-ply analysis says take (.006), Roller+ and Roller++ agree on pass (.015 amd .018), and a full 1296 rollout has the take within the margin of error.

So, this is a close one folks. If I hadn't misplayed the 32 it would make a nice reference position. Here's the result of an extended rollout, showing it to be a narrow take. Note that the difference between gnu 2-ply analysis (which said it was a clear take) and this rollout (which says close pass) is the GWC chance difference of 62.0 vs 61.7 and the gammon rate of 18.36 vs 17.3. That extra 1% gammons makes a big difference, and I don't know that I'll ever be able to estimate that precisely OTB no matter how many positions I memorize.

 

Analyzed in Rollout 

No double 

  Player Winning Chances:   62.35% (G:17.28% B:0.80%) 

  Opponent Winning Chances: 37.65% (G:8.60% B:0.88%) 

Double/Take 

  Player Winning Chances:   62.07% (G:18.36% B:1.16%) 

  Opponent Winning Chances: 37.93% (G:8.77% B:0.91%) 

Cubeless Equities: No Double=+0.272, Double=+1.015 

Cubeful Equities: 

       No double:     +0.790 (-0.210) 

       Double/Take:   +1.015 (+0.015) 

       Double/Pass:   +1.000 

Best Cube action: Double / Pass 

Rollout: 

  2592 Games rolled with Variance Reduction. 

  Moves: 3-ply, cube decisions: XG Roller 

  Confidence No Double: ± 0.008 (+0.782..+0.798) 

  Confidence Double:    ± 0.012 (+1.003..+1.027) 

  Double Decision confidence: 100.0% 

  Take Decision confidence: 99.4% 

  Duration: 1 hour 04 minutes 

eXtreme Gammon Version: 2.03, MET: Kazaross XG2 

Note: I'm still figuring out XG and haven't discovered how to post the nice html tables yet.

Messages In This Thread

• 2-away 4-away cube decision
  ah_clem -- Friday, 13 April 2012, at 4:32 p.m.
  • 53P-66B-32E-C-4a2a
    Stick -- Friday, 13 April 2012, at 4:40 p.m.
    • 53P-66B-32W-C.–4–2
      Nack Ballard -- Sunday, 15 April 2012, at 3:14 a.m.
  • 2-away 4-away cube decision
    David Rennie -- Saturday, 14 April 2012, at 12:03 a.m.
  • 2-away 4-away cube decision
    Timothy Chow -- Saturday, 14 April 2012, at 1:42 a.m.
    • 2-away 4-away cube decision
      Stick -- Saturday, 14 April 2012, at 5:33 p.m.
      • 2-away 4-away cube decision
        Timothy Chow -- Saturday, 14 April 2012, at 6:41 p.m.
        • 2-away 4-away cube decision
          Stick -- Saturday, 14 April 2012, at 7:45 p.m.
  • 2-away 4-away cube decision
    neilkaz -- Saturday, 14 April 2012, at 11:57 p.m.
  • 2-away 4-away cube decision
    ah_clem -- Monday, 16 April 2012, at 2:13 p.m.
    • 53P-66B-32W-C.–4–2 Rollout
      ah_clem -- Monday, 16 April 2012, at 2:34 p.m.