BGonline.org Forums

5pt and 8pt Variants

Posted By: Nack Ballard
Date: Wednesday, 9 September 2015, at 1:15 a.m.

In Response To: Run or build? - XGR++ plus VARIANTS - insight needed (Karol Szczerek)

		White is Player 2 score: 0 pip: 138 Unlimited Game Jacoby Beaver pip: 154 score: 0 Blue is Player 1
XGID=-b--B-D-C--AbC---c-cbc--B-:0:0:1:62:0:0:3:0:10
4pt Position (original)

1.	XG Roller++	13/5	eq: -0.529
	Player: Opponent: 38.55% (G:10.37% B:0.45%) 61.45% (G:25.32% B:2.70%)
2.	XG Roller++	13/7 11/9	eq: -0.612 (-0.083)
	Player: Opponent: 37.37% (G:10.12% B:0.47%) 62.63% (G:27.03% B:2.99%)
3.	XG Roller++	24/16	eq: -0.630 (-0.101)
	Player: Opponent: 38.62% (G:9.94% B:0.59%) 61.38% (G:31.74% B:1.37%)

		White is Player 2 score: 0 pip: 139 Unlimited Game Jacoby Beaver pip: 154 score: 0 Blue is Player 1
XGID=-b--B-D-C--AbC---c-ccb--B-:0:0:1:62:0:0:3:0:10
5pt Variant

1.	Rollout1	13/5	eq: -0.561
	Player: Opponent: 37.31% (G:10.91% B:0.49%) 62.69% (G:25.56% B:2.86%) Conf.: กำ 0.010 (-0.571...-0.551) - [99.4%] Duration: 1 hour 31 minutes
2.	Rollout1	24/16	eq: -0.579 (-0.018)
	Player: Opponent: 38.93% (G:10.50% B:0.64%) 61.07% (G:31.86% B:1.41%) Conf.: กำ 0.009 (-0.588...-0.569) - [0.6%] Duration: 1 hour 32 minutes
3.	Rollout1	24/18 13/11	eq: -0.630 (-0.069)
	Player: Opponent: 38.54% (G:10.80% B:0.72%) 61.46% (G:34.44% B:1.14%) Conf.: กำ 0.010 (-0.640...-0.619) - [0.0%] Duration: 1 hour 16 minutes
1 5184 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller

 

When I saw your initial post (see first position above), running looked wrong, but the first alternative position I considered was the "5pt variant" (what you call "Variant A"). I noticed that if Blue runs, most of White's 3s (including 21) play less efficiently than in the original position.

Put another way, in the 4pt position, R diversifies White's 4s and 3s, whereas in the 5pt variant R duplicates 4s. As the rollout above (second position) reveals (as you noted too, comparing eval to eval), R is much, much closer (i.e., when the spare is on the 5pt rather than the 4pt), but R still falls short of $ (by .018).

On my next try, I leapt past the breakeven with what I call the "8pt variant." A rollout is included below.

Nack

		White is Player 2 score: 0 pip: 142 Unlimited Game Jacoby Beaver pip: 154 score: 0 Blue is Player 1
XGID=-b--B-D-C--AbC---d-cbb--B-:0:0:1:26:0:0:3:0:10
8pt Variant

1.	Rollout1	24/16	eq: -0.466
	Player: Opponent: 40.68% (G:11.17% B:0.66%) 59.32% (G:26.71% B:1.41%) Conf.: กำ 0.007 (-0.473...-0.459) - [100.0%] Duration: 1 hour 49 minutes
2.	Rollout1	24/18 13/11	eq: -0.515 (-0.048)
	Player: Opponent: 39.93% (G:11.14% B:0.69%) 60.07% (G:27.34% B:1.14%) Conf.: กำ 0.007 (-0.522...-0.507) - [0.0%] Duration: 1 hour 40 minutes
3.	Rollout1	13/5	eq: -0.538 (-0.071)
	Player: Opponent: 37.47% (G:11.22% B:0.51%) 62.53% (G:23.81% B:2.23%) Conf.: กำ 0.010 (-0.548...-0.528) - [0.0%] Duration: 1 hour 28 minutes
1 5184 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller

Messages In This Thread

• Run or build?
  Karol Szczerek -- Saturday, 5 September 2015, at 10:42 a.m.
  • Run or build?
    Wolfgang -- Saturday, 5 September 2015, at 3:47 p.m.
    • Run or build?
      Chase -- Saturday, 5 September 2015, at 5:13 p.m.
      • Run or build?
        Timothy Chow -- Saturday, 5 September 2015, at 9:01 p.m.
  • Run or build? - XGR++ plus VARIANTS - insight needed
    Karol Szczerek -- Tuesday, 8 September 2015, at 10:17 p.m.
    • Run or build? - XGR++ plus VARIANTS - insight needed
      Timothy Chow -- Wednesday, 9 September 2015, at 12:28 a.m.
    • 5pt and 8pt Variants
      Nack Ballard -- Wednesday, 9 September 2015, at 1:15 a.m.