BGonline.org Forums

Midgame 64 - Rollout and VARIANTS

Posted By: Nack Ballard
Date: Monday, 6 November 2017, at 4:46 a.m.

In Response To: Midgame 64 - RO (Karol Szczerek)

		White is Player 2 score: 0 pip: 169 Unlimited Game pip: 131 score: 0 Blue is Player 1
XGID=--Ba-cD-C---bEA--cae------:0:0:1:64:0:0:0:0:10
Original position

1.	Rollout1	14/4	eq: +0,626
	Player: Opponent: 65,66% (G:14,62% B:0,41%) 34,34% (G:6,71% B:0,22%) Conf.: ± 0,008 (+0,618...+0,634) - [100,0%] Duration: 47 minutes 15 seconds
2.	Rollout1	13/3*	eq: +0,596 (-0,030)
	Player: Opponent: 63,82% (G:17,61% B:0,53%) 36,18% (G:7,82% B:0,27%) Conf.: ± 0,007 (+0,589...+0,604) - [0,0%] Duration: 42 minutes 54 seconds
3.	Rollout2	14/8 6/2	eq: +0,562 (-0,064)
	Player: Opponent: 63,72% (G:15,83% B:0,39%) 36,28% (G:7,76% B:0,28%) Conf.: ± 0,010 (+0,552...+0,571) - [0,0%] Duration: 26 minutes 44 seconds
1 5184 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller 2 2592 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller

Karol's rollout of his original position is copied above.

I rolled out three related positions, each of which advances two of White's 6pt spares to make an inside point:

My result for White's 5pt is that N (Near, 14/8 6/2) beats $ (Slot, 14/4) by .030 (5k). Karol independently rolled this out (5k) and garnered a margin of .023. (Combining our rollouts yields a 10k margin of .026 or .027). [Stick mentioned the possibility of this variant as well though didn't roll it out.]

In search of the breakeven point, I made a lucky guess when I tried White's 3pt next, as you can see in Variant 1 below.

		White is Player 2 score: 0 pip: 163 Unlimited Game Jacoby Beaver pip: 131 score: 0 Blue is Player 1
XGID=--Ba-cD-C---bEA--cac--b---:0:0:1:64:0:0:3:0:10
Variant 1 (Compared to the original, White has played 6/3(2))

1.	Rollout1	14/4	eq: +0.476
	Player: Opponent: 62.34% (G:12.70% B:0.36%) 37.66% (G:8.39% B:0.27%) Conf.: ± 0.004 (+0.472...+0.480) - [78.7%] Duration: 1 hour 58 minutes
2.	Rollout1	14/8 6/2	eq: +0.474 (-0.002)
	Player: Opponent: 61.83% (G:13.45% B:0.30%) 38.17% (G:8.62% B:0.27%) Conf.: ± 0.004 (+0.469...+0.478) - [21.3%] Duration: 1 hour 56 minutes
3.	Rollout2	13/3*	eq: +0.378 (-0.098)
	Player: Opponent: 59.38% (G:14.98% B:0.46%) 40.62% (G:10.32% B:0.37%) Conf.: ± 0.014 (+0.364...+0.392) - [0.0%] Duration: 10 minutes 29 seconds
1 15552 Games rolled with Variance Reduction. Dice Seed: 6286528 Moves: 3-ply, cube decisions: XG Roller 2 1296 Games rolled with Variance Reduction. Dice Seed: 6286528 Moves: 3-ply, cube decisions: XG Roller

In the in-between position with White owning her 4pt, N beats $ by .018 (rollout of 5k not shown). As you might expect, this margin is between, though not halfway between, the 5pt margin of .027 and the 3pt margin of –.002.

So far, this post shows a diagram of one position in which $ is the best play, and one in which N is the best play. Is there a related position where H (Hit, 13/3*) is the best play? Yes, for example see Variant 2a below:

		White is Player 2 score: 1 pip: 165 2 point match Crawford pip: 131 score: 0 Blue is Player 1
XGID=--Ba-cD-C---bEA--bae-a----:0:0:1:64:0:1:1:2:10
Variant 2a (Compared to the original, White has slotted 8/4

1.	Rollout1	13/3*	eq: +0.576
	Player: Opponent: 67.54% (G:11.27% B:0.59%) 32.46% (G:6.13% B:0.51%) Conf.: ± 0.004 (+0.572...+0.580) - [100.0%] Duration: 16 minutes 41 seconds
2.	Rollout1	14/4	eq: +0.527 (-0.049)
	Player: Opponent: 68.06% (G:8.31% B:0.33%) 31.94% (G:5.17% B:0.49%) Conf.: ± 0.004 (+0.524...+0.531) - [0.0%] Duration: 15 minutes 41 seconds
3.	Rollout2	14/8 6/2	eq: +0.499 (-0.077)
	Player: Opponent: 65.31% (G:9.64% B:0.60%) 34.69% (G:6.73% B:0.56%) Conf.: ± 0.006 (+0.493...+0.505) - [0.0%] Duration: 8 minutes 16 seconds
1 2592 Games rolled with Variance Reduction. Dice Seed: 73501684 Moves: 3-ply, cube decisions: XG Roller 2 1296 Games rolled with Variance Reduction. Dice Seed: 73501684 Moves: 3-ply, cube decisions: XG Roller

However, I confess that I stacked the deck some. Suppose, instead of slotting with 8/4, White slots with 6/4. This emphasizes 2 duplication less (while still maintaining the 3 duplication) and makes White's distribution less awkward. Suddenly, Blue's H is demoted to a tie with $ (the latter actually boasting a tiny lead), as you can see in Variant 2b below.

Nack

		White is Player 2 score: 0 pip: 167 Unlimited Game Jacoby Beaver pip: 131 score: 0 Blue is Player 1
XGID=--Ba-cD-C---bEA--cad-a----:0:0:1:64:0:0:3:0:10
Variant 2b (Compared to the original, White has slotted 6/4)

1.	Rollout1	14/4	eq: +0.525
	Player: Opponent: 63.20% (G:14.02% B:0.38%) 36.80% (G:7.69% B:0.26%) Conf.: ± 0.003 (+0.522...+0.528) - [82.3%] Duration: 4 hours 02 minutes
2.	Rollout1	13/3*	eq: +0.523 (-0.002)
	Player: Opponent: 61.94% (G:18.66% B:0.69%) 38.06% (G:8.67% B:0.32%) Conf.: ± 0.004 (+0.519...+0.527) - [17.7%] Duration: 4 hours 09 minutes
3.	Rollout2	14/8 6/2	eq: +0.489 (-0.036)
	Player: Opponent: 61.88% (G:14.92% B:0.30%) 38.12% (G:8.34% B:0.28%) Conf.: ± 0.008 (+0.481...+0.496) - [0.0%] Duration: 48 minutes 08 seconds
1 31104 Games rolled with Variance Reduction. Dice Seed: 13417452 Moves: 3-ply, cube decisions: XG Roller 2 5184 Games rolled with Variance Reduction. Dice Seed: 13417452 Moves: 3-ply, cube decisions: XG Roller

Messages In This Thread

• Midgame 64
  Karol Szczerek -- Friday, 3 November 2017, at 10:20 p.m.
  • Midgame 64
    Stick -- Sunday, 5 November 2017, at 3:26 a.m.
    • Midgame 64
      neilkaz -- Sunday, 5 November 2017, at 3:01 p.m.
  • Midgame 64 - RO
    Karol Szczerek -- Sunday, 5 November 2017, at 9:30 p.m.
    • Midgame 64 - Rollout and VARIANTS
      Nack Ballard -- Monday, 6 November 2017, at 4:46 a.m.