Open Side Menu Go to the Top
Register
the big thread about cube decisions the big thread about cube decisions

03-12-2015 , 09:06 AM
Good heavens! Miles off. I would have accepted with the blissfully ignorant approach that it had been a thin double, and I was fighting fit.
the big thread about cube decisions Quote
03-12-2015 , 10:07 AM
Black doubles because he has plenty of market losers and there's a ton of volatility. White takes because he's ahead in the race and has a better board.

What's puzzling me is trying to figure out what White's last roll and play were. I can't see how he got to this position without making a pretty big blunder.
the big thread about cube decisions Quote
03-12-2015 , 12:02 PM
White could have been on the bar and rolled a 3-1.
the big thread about cube decisions Quote
03-12-2015 , 01:43 PM
Quote:
Originally Posted by _Z_
White could have been on the bar and rolled a 3-1.
Good point. That works.
the big thread about cube decisions Quote
03-12-2015 , 04:58 PM
Position ID: bJ6RAUTYtkFANA Match ID: UgkAAAAAAAAE

White - Pips 138

Black - Pips 164
Black on roll. Cube action?
Created with www.BGdiagram.com
the big thread about cube decisions Quote
03-12-2015 , 06:19 PM
Double/Take.
the big thread about cube decisions Quote
03-13-2015 , 01:01 AM
This looks to me like it borders on too good to double, apart from having so many men back. My approach with "too good to double" situations is to double anyway, unless it is glaringly obvious, since I seem to get them wrong so often and miss an opportunity.

So I would double here, and I'm not sure it's a take.
the big thread about cube decisions Quote
03-13-2015 , 02:17 AM
Bleep, you are cute. How can it border on too good to double and still be a take.
the big thread about cube decisions Quote
03-13-2015 , 03:05 AM
I said I wasn't sure it's a take, meaning a drop.

On a side note, that's the nicest thing anyone has said to me today, but it's still early here.
the big thread about cube decisions Quote
03-13-2015 , 03:49 AM
I would say double. I would definitely take as white as I'm assuming that white has a chance of holding black back long enough to escape. Black has his trap now, but white could develop a reasonable trap soon and hope to throw a six before it collapses.
the big thread about cube decisions Quote
03-13-2015 , 05:26 AM
I'll go with too-good/pass.

Fully half of Black's rolls hit another checker, and many of the ones that miss will make the 20pt. Even when Black does neither of those things, White is still dancing a quarter of the time. If White does not dance, his 4s and 5s could be awkward.

Mike
the big thread about cube decisions Quote
03-13-2015 , 09:41 AM
Tough position.

Too good? No way. Black trails by 26 in the race and he still has four checkers in White's home board. His 3-point board doesn't rate to get better quickly. He can't try to turn the game into a blitz because White's board is too strong, and a stripped 5-prime in a long game isn't as good as it looks. Playing on for the gammon must be a blunder.

Good enough to double? Sure. Two men behind a 5-prime with a 50% shot at hitting a third looks good enough. White has points but they're not consecutive so Black's numbers are all working. He should be able to wiggle out.

Take or drop? White's up 26 and he's 75% to enter every turn, so he should be in while Black is still wiggling out. I'll take and look forward to shipping a scary-looking 16-cube if things turn around. Those can be hard to take.
the big thread about cube decisions Quote
03-13-2015 , 09:45 AM
Thanks for the explanation.

Code:
Cube analysis
Rollout cubeless equity  +0,591

Cubeful equities:
1. Double, take         +0,979
2. Double, pass         +1,000  ( +0,021)
3. No double            +0,863  ( -0,115)
Proper cube action: Redouble, take
the big thread about cube decisions Quote
03-13-2015 , 05:19 PM
Position ID: bJ6RAUTYtkFANA Match ID: UgkAAAAAAAAE

White - Pips 138

Black - Pips 164
Black on roll. Cube action?

XG has this as a small pass, which makes sense to me. In half the games, White will start out with 2 checkers on the bar, and 3 behind a five-prime. That is not a comfortable beginning.

I see now that Black has too much work to do for this to be too good. Extracting the 4 checkers he has in White’s board will take some time. If you move Black’s anchor on the 24pt up to the 18pt, then this becomes too good.

5k XG Rollout
Unlimited game, no Jacoby, no beavers
Code:
XGID=a--aBBBBB---bAa--b-dAbbAB-:2:1:1:00:0:0:0:0:10

X:Player 1   O:Player 2
Score is X:0 O:0. Unlimited Game
Pip count  X: 164  O: 138 X-O: 0-0
Cube: 4, X own cube
X on roll, cube action

Analyzed in Rollout
No redouble
  Player Winning Chances:   71.14% (G:27.59% B:1.60%)
  Opponent Winning Chances: 28.86% (G:9.13% B:0.68%)
Redouble/Take
  Player Winning Chances:   71.39% (G:28.12% B:1.69%)
  Opponent Winning Chances: 28.61% (G:9.17% B:0.68%)

Cubeless Equities: No Double=+0.617, Double=+1.255

Cubeful Equities:
       No redouble:     +0.894 (-0.106)
       Redouble/Take:   +1.045 (+0.045)
       Redouble/Pass:   +1.000

Best Cube action: Redouble / Pass

Rollout:
  5184 Games rolled with Variance Reduction.
  Dice Seed: 80389452
  Moves: 3-ply, cube decisions: XG Roller
  Confidence No Double: ± 0.006 (+0.888..+0.900)
  Confidence Double:    ± 0.011 (+1.034..+1.056)

  Double Decision confidence: 100.0%
  Take Decision confidence: 100.0%

eXtreme Gammon Version: 2.10
Mike
the big thread about cube decisions Quote
03-13-2015 , 06:04 PM
With equities that are very close to those of the XG rollout above, GnuBg also has this as a small pass. Evidently, Yogiman’s 147-trial was too short to detect this. Either that, or you need a full rollout instead of a trunc to pick up the difference.

1k GnuBg Rollout
Unlimited game, no Jacoby, no beavers

Code:
Cube analysis
Rollout cubeless equity  +0.605

Cubeful equities:
1. Double, pass         +1.000
2. Double, take         +1.039  ( +0.039)
3. No double            +0.886  ( -0.114)

Proper cube action: Redouble, pass 

Rollout details:
Player Mike.TutorMode owns 4-cube:
  0.707 0.277 0.016 - 0.293 0.095 0.006 CL  +0.605 CF  +0.886
 [0.001 0.001 0.000 - 0.001 0.001 0.000 CL   0.003 CF   0.007]
Player GnuBg owns 8-cube:
  0.713 0.281 0.016 - 0.287 0.092 0.006 CL  +1.254 CF  +1.039
 [0.002 0.002 0.001 - 0.002 0.001 0.000 CL   0.008 CF   0.012]
Full cubeful rollout with var.redn.
1296 games, Mersenne Twister dice gen. with seed 740666243 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]
Mike
the big thread about cube decisions Quote
03-18-2015 , 03:52 AM
Position ID: 3G4jQEDbvQEBEA Match ID: UQkAAAAAAAAE

White - Pips 111

Black - Pips 76
Black on roll. Cube action?
Created with www.BGdiagram.com
the big thread about cube decisions Quote
03-18-2015 , 05:52 AM
Double/take. Both guys look to be in trouble without the right dice, so it's volatile. I'd almost be tempted to drop as white because of all the loose blots lying around, but I think I can find a take here. Would be nervous, as always, in situations like this, where an error is normally a major blunder!
the big thread about cube decisions Quote
03-18-2015 , 07:25 AM
OTB I would double. But only because I would be hoping white would drop. Which is probably a terrible reason. My expectation here would be that it could turn very bad for black so I would be trying to get white to drop now while I'm still ok. If I was playing a guy i knew took everything, I would not double and let the dice decide, purely because I have no clue what the right play is!
the big thread about cube decisions Quote
03-18-2015 , 07:26 AM
By the way, as white, I would normally drop if doubled here.
the big thread about cube decisions Quote
03-18-2015 , 02:37 PM
Quote:
Originally Posted by Kamba
OTB I would double. But only because I would be hoping white would drop. Which is probably a terrible reason.
Usually you dont have to offer a cube youre afraid to Be taken. But here black has any 6 as a market loser and probably two rolls to get them in case of bad rolling. Its a double imo even because its not clear if on white shoes is a rake or a drop. With white id drop otb because of the blot but i think that its a computer take because of high point open so final answer is double take
the big thread about cube decisions Quote
03-18-2015 , 11:47 PM
Double/Drop.
the big thread about cube decisions Quote
03-19-2015 , 06:42 AM
Black has 24 rolls that either hit or escape. When White does not come back with a strong reply, those are market-losers. If Black can hit a second White checker, he may well pick up a third. He is probably netting a gammon victory in something like 20% of all games.

Black has 15 hitters. On the 26 non-hitters, White will have a 30% chance to enter, so that's about 8. Something like half of them will give White a return hit. Those do not sound like good odds to me.

I pass.

Mike
the big thread about cube decisions Quote
03-20-2015 , 10:34 AM
This is again a borderline case, but I think as humans we should better pass, in order not to get into a habit of detrimentally loose takes.
Code:
Cube analysis 
Rollout cubeless equity  +0,619 

Cubeful equities: 
1. Double, pass         +1,000 
2. Double, take         +1,014  ( +0,014) 
3. No double            +0,762  ( -0,238) 
Proper cube action: Redouble, pass
the big thread about cube decisions Quote
03-21-2015 , 04:10 AM
5k XG Rollout – Unlimited game, no Jacoby, no beavers
Code:
XGID=aBBCDB-a----A--a--bbcbcA--:1:1:1:00:0:0:0:0:10

X:Player 1   O:Player 2
Score is X:0 O:0. Unlimited Game
Pip count  X: 76  O: 111 X-O: 0-0
Cube: 2, X own cube
X on roll, cube action

Analyzed in Rollout
No redouble
  Player Winning Chances:   66.72% (G:30.32% B:0.52%)
  Opponent Winning Chances: 33.28% (G:3.12% B:0.09%)
Redouble/Take
  Player Winning Chances:   67.51% (G:29.91% B:0.48%)
  Opponent Winning Chances: 32.49% (G:3.27% B:0.09%)

Cubeless Equities: No Double=+0.611, Double=+1.241

Cubeful Equities:
       No redouble:     +0.760 (-0.240)
       Redouble/Take:   +1.011 (+0.011)
       Redouble/Pass:   +1.000

Best Cube action: Redouble / Pass

Rollout:
  5184 Games rolled with Variance Reduction.
  Dice Seed: 9333068
  Moves: 3-ply, cube decisions: XG Roller
  Confidence No Double: ± 0.003 (+0.756..+0.763)
  Confidence Double:    ± 0.005 (+1.006..+1.015)

  Double Decision confidence: 100.0%
  Take Decision confidence: 100.0%

  Duration: 7 minutes 58 seconds

Rollout by Taper_Mike
2015-Mar-20
eXtreme Gammon Version: 2.10.199.2658
Cube Decisions 009.xgp
Mike
the big thread about cube decisions Quote
03-27-2015 , 11:46 AM
Position ID: 2zgLDBCz3QCCAw Match ID: cAkAAAAAAAAE

White - Pips 108

Black - Pips 118
Black on roll. Cube action?
Created with www.BGdiagram.com
the big thread about cube decisions Quote

      
m