Game 5, move 1 : Forum 6-2 - roadkillbooks CUBE?

[diane]

Forum, blue, to move 6-2

4NvgATDgc/ABMA:cAmrACAAGAAA

[ah_clem]

(click to show/hide)

Slotting our 5 pt or bringing down two builders are "gammon go" plays.  I don't think they're the right approach here.

The match score is more "gammon save", so I say escape a backman.  24/16.

[diane]

Looks like we have a choice...I will move this along.

[diane]

Actually, I will throw out more cubing opportunities, with things as they are  ..Therefore - roadkillbooks - do you fancy cubing at this point?

[ah_clem]

gnubg rollout

(click to show/hide)

No surprises here. 


    1. Rollout          24/16                        Eq.:  -0.446
       0.448 0.147 0.018 - 0.552 0.169 0.018 CL  -0.259 CF  -0.446
      [0.002 0.005 0.003 - 0.002 0.002 0.002 CL   0.005 CF   0.014]

    2. Rollout          24/22 13/7                   Eq.:  -0.549 ( -0.103)
       0.430 0.161 0.019 - 0.570 0.171 0.027 CL  -0.297 CF  -0.549
      [0.002 0.005 0.002 - 0.002 0.002 0.003 CL   0.006 CF   0.016]
       
    3. Rollout          13/5                         Eq.:  -0.553 ( -0.106)
       0.428 0.155 0.019 - 0.572 0.167 0.039 CL  -0.298 CF  -0.553
      [0.002 0.005 0.003 - 0.002 0.002 0.003 CL   0.006 CF   0.016]

    4. Rollout          13/11 13/7                   Eq.:  -0.602 ( -0.156)
       0.422 0.152 0.019 - 0.578 0.174 0.033 CL  -0.316 CF  -0.602
      [0.003 0.005 0.003 - 0.003 0.003 0.003 CL   0.007 CF   0.019]
       
    5. Rollout          13/7 6/4                     Eq.:  -0.683 ( -0.236)
       0.411 0.141 0.018 - 0.589 0.182 0.047 CL  -0.345 CF  -0.683
      [0.004 0.005 0.002 - 0.004 0.003 0.004 CL   0.009 CF   0.025]
        Full cubeful rollout with var.redn.
        1296 games, Mersenne Twister dice gen. with seed 865264095 and quasi-random dice
        Play: 0-ply cubeful prune [expert]
        Cube: 0-ply cubeful prune [expert]

[roadkillbooks]

NO DOUBLE

[ah_clem]

gnubg results for the cube decision

(click to show/hide)

Interesting results. analysis disagrees with rollout.  I'll believe the rollout - it doesn't look like a double.

Cube analysis
0-ply cubeless equity  +0.169 (Money:  +0.184)
  0.561 0.166 0.005 - 0.439 0.105 0.004
Cubeful equities:
1. Double, take         +0.488
2. Double, pass         +1.000  ( +0.512)
3. No double            +0.450  ( -0.038)
Proper cube action: Double, take


Cube analysis
Rollout cubeless equity  +0.236 (Money:  +0.138)

Cubeful equities:
1. No double            +0.499
2. Double, pass         +1.000  ( +0.501)
3. Double, take         +0.446  ( -0.053)
Proper cube action: No double, take (9.6%)
Rollout details:
Centered 1-cube:
  0.553 0.171 0.015 - 0.447 0.142 0.012 CL  +0.236 CF  +0.499
 [0.002 0.002 0.002 - 0.002 0.004 0.003 CL   0.005 CF   0.013]
Player You owns 2-cube:
  0.552 0.169 0.018 - 0.448 0.147 0.018 CL  +0.446 CF  +0.446
 [0.002 0.002 0.002 - 0.002 0.005 0.003 CL   0.014 CF   0.014]
Full cubeful rollout with var.redn.
1296 games, Mersenne Twister dice gen. with seed 865264095 and quasi-random dice
Play: 0-ply cubeful prune [expert]
Cube: 0-ply cubeful prune [expert]

[dorbel]

Gnu 0-ply analysis, particularly of cube actions, is worthless. 2-ply is a must.
Not that it's very important clem, but if you publish rollout equity of our move before rkb makes his cube decision, it's a bit helpful isn't it? After he has doubled or not might be better!

[ah_clem]

Gnu 0-ply analysis, particularly of cube actions, is worthless. 2-ply is a must.
Not that it's very important clem, but if you publish rollout equity of our move before rkb makes his cube decision, it's a bit helpful isn't it? After he has doubled or not might be better!

Gnu 0-ply analysis, particularly of cube actions, is worthless?  How so?  Not trying to be argumentative, I'm just curious.  I've been practicing against gnubg, and analyzing all my FIBS matches with it using 0-Ply settings, chosen so that the "analyze match" process finishes before I've forgotten what I was thinking during the match.  Perhaps I should use a higher setting?

As for publishing the rollout, it's behind a spoiler with the warning that it contains a rollout.  I assume that rkb is a good enough sport not to look.  Of course, it might influence *our* take decision, so waiting should be the default behavior and I'll do that in the future.  In this case the take was so clear (to me anyway) that I didn't see any harm in posting it.

[adrian]

The difference between 0ply and 2ply is huge sometimes, see for yourself clem. Try 2ply analysys and rollouts and compare them with the 0ply.