Quote:
Originally Posted by networth
Has the second roll responses been updated anywhere online? I'm pretty sure 32W in response to 62S and 63S has been changed but I have refused to give it up so far.
Diagram 1. Black to Play 3-2
The “splot” is very much alive! For those unfamiliar with the term, a splot is a play where you split and slot on the same move. Nack Ballard calls such moves “wild,” hence the use of the letter W to describe them. In Diagram 1, where White opened with 62-split, and now Black must play a 32, w = 24/21, 6/4.
The following 46k, 4-ply rollout was completed by Dmitry Obukhov in September, 2013, too late for inclusion in the latest version of the XG Opening Book. It shows that the top three plays are almost tied, with w = 24/21, 6/4 trailing the leader by only 0.006 points of equity.
Code:
XGID=-a----EaC---dEa--c-e----B-:0:0:1:32:0:0:3:0:10
X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game, Jacoby Beaver
Pip count X: 167 O: 159 X-O: 0-0
Cube: 1
X to play 32
1. Rolloutą 24/22 24/21 eq:-0.2499
Player: 43.97% (G:9.82% B:0.41%)
Opponent: 56.03% (G:14.36% B:0.56%)
Confidence: ±0.0021 (-0.2520..-0.2478) - [100.0%]
Duration: 1 day 14 hours 33 minutes
2. Rolloutą 24/21 6/4 eq:-0.2559 (-0.0060)
Player: 44.11% (G:10.69% B:0.43%)
Opponent: 55.89% (G:15.41% B:0.75%)
Confidence: ±0.0021 (-0.2580..-0.2538) - [0.0%]
Duration: 1 day 16 hours 44 minutes
3. Rolloutą 24/21 13/11 eq:-0.2578 (-0.0079)
Player: 43.91% (G:10.88% B:0.43%)
Opponent: 56.09% (G:15.17% B:0.74%)
Confidence: ±0.0020 (-0.2598..-0.2558) - [0.0%]
Duration: 1 day 13 hours 59 minutes
ą 46656 Games rolled with Variance Reduction.
Dice Seed: 96
Moves: 4-ply, cube decisions: XG Roller
eXtreme Gammon Version: 2.10
Diagram 2. Black to Play 3-2
In Diagram 2, White opened with 63-split, playing 24/18, 13/10. Now Black must play a 32 in reply. In 2011 and 2012, Nack Ballard headed up a team of rollateers that made an intensive investigation of this position. Some of the longest 4-ply rollouts ever made in XG resulted in a near tie for the top three plays. After 40 days and 228,000 trials, these were the results:
1st place: S = Split = 24/21, 13/11
2nd place: Z = Reverse split = 24/22, 13/10, trailing by 0.0018 points of equity (1.8 millipoints)
3rd place: w = wild = 24/21, 64, trailing by 0.0025 points of equity (2.5 millipoints)
4th place: U = Up = 24/22, 24/21, trailing by 0.0065 points of equity ( 6.5 millipoints)
You can
see this amazing rollout at BgOnline.
The XG Opening Book v. 3.70 uses a 3-ply rollout with 186,624 trials. It yields similar results:
Code:
XGID=-a----EaC---dE-a-c-e----B-:0:0:1:32:0:0:3:0:10
X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game, Jacoby Beaver
Pip count X: 167 O: 158 X-O: 0-0
Cube: 1
X to play 32
1. Booką 24/21 13/11 eq:-0.2731
Player: 43.69% (G:10.82% B:0.44%)
Opponent: 56.31% (G:15.87% B:0.81%)
Confidence: ±0.0013 (-0.2744..-0.2718) - [99.0%]
2. Booką 24/21 6/4 eq:-0.2754 (-0.0023)
Player: 43.92% (G:10.83% B:0.46%)
Opponent: 56.08% (G:16.92% B:0.92%)
Confidence: ±0.0014 (-0.2768..-0.2740) - [0.9%]
3. Booką 24/22 13/10 eq:-0.2762 (-0.0031)
Player: 43.38% (G:10.52% B:0.41%)
Opponent: 56.62% (G:15.34% B:0.70%)
Confidence: ±0.0013 (-0.2775..-0.2749) - [0.0%]
4. Booką 24/22 24/21 eq:-0.2763 (-0.0032)
Player: 43.40% (G:9.74% B:0.39%)
Opponent: 56.60% (G:14.82% B:0.59%)
Confidence: ±0.0013 (-0.2775..-0.2750) - [0.0%]
ąGenerated by Christopher Yep on 5/3/2012 using eXtreme Gammon 2.00
186624 Games rolled with Variance Reduction.
Dice Seed: 31206198
Moves and cube decisions: 3-ply
These are just two of the many plays where splotting is correct (or at least tied).
Mike