| 1. | Rollout1 | Bar/22 24/22 | eq: +0.047 |
| Player:
Opponent: | 50.03% (G:17.19% B:0.68%)
49.97% (G:14.01% B:0.64%) | Conf.: ± 0.007 (+0.040...+0.054) - [100.0%]
Duration: 1 hour 17 minutes |
|
| 2. | Rollout1 | Bar/23* 11/8 | eq: -0.067 (-0.114) |
| Player:
Opponent: | 47.69% (G:17.03% B:1.51%)
52.31% (G:16.56% B:0.77%) | Conf.: ± 0.008 (-0.075...-0.060) - [0.0%]
Duration: 1 hour 29 minutes |
|
| 3. | Rollout1 | Bar/23* 13/10 | eq: -0.094 (-0.140) |
| Player:
Opponent: | 46.85% (G:18.61% B:1.89%)
53.15% (G:18.64% B:1.34%) | Conf.: ± 0.009 (-0.102...-0.085) - [0.0%]
Duration: 1 hour 25 minutes |
|
| |
1 5184 Games rolled with Variance Reduction.
Dice Seed: 64026929
Moves: 3-ply, cube decisions: XG Roller
|
| 1. | Rollout1 | Bar/22 24/22 | eq: +0.208 |
| Player:
Opponent: | 53.96% (G:18.35% B:0.81%)
46.04% (G:12.46% B:0.54%) | Conf.: ± 0.007 (+0.201...+0.215) - [100.0%]
Duration: 1 hour 19 minutes |
|
| 2. | Rollout1 | Bar/23* 13/10 | eq: +0.147 (-0.061) |
| Player:
Opponent: | 52.26% (G:22.27% B:2.55%)
47.74% (G:17.04% B:1.31%) | Conf.: ± 0.010 (+0.137...+0.156) - [0.0%]
Duration: 1 hour 24 minutes |
|
| 3. | Rollout1 | Bar/23* 11/8 | eq: +0.111 (-0.097) |
| Player:
Opponent: | 51.79% (G:19.74% B:1.77%)
48.21% (G:15.15% B:0.74%) | Conf.: ± 0.009 (+0.102...+0.120) - [0.0%]
Duration: 1 hour 29 minutes |
|
| |
1 5184 Games rolled with Variance Reduction.
Dice Seed: 64026929
Moves: 3-ply, cube decisions: XG Roller
|
The purpose of this post is to demonstrate that Blue's decision to hit or anchor is affected by the strength of his own board. If White ends up creating a middle or high anchor, the reward from Blue's hitting will be substantially diminished. In short, the higher the quality of Blue's points, the more incentive he has to hit.
There are four positions rolled out. If your window is wide enough, you can see two side-by-side above, and two side-by-side below (requires less scrolling).
The first position (of the four) is the one leobueno posted originally here, then he posted a (very short) GnuBG rollout here. The second position (a variant I suggested here), which upgrades Blue's 2pt to his 3pt, gives Blue a closer hit/anchor decision (though, admittedly, not as close as I would have guessed). The third position gives Blue an additional upgrade, 4pt to 5pt, which gives us the "tipping point" -- anchoring and both hitting moves all become very close. The fourth position (suggested by Stick here) makes a final upgrade, 3pt to 4pt, where hitting is clearly better than anchoring.
Below is a summary table of the results. The three-digit number in the left column denotes the three inside points that Blue owns (e.g., "642" means he owns the 6pt, 4pt and 2pt). Reading down, as Blue's board strengthens, his hitting plays (B and Z) are promoted and his anchoring play (@) demoted.
........642 ....[@ h114 H140] "<=5
........643 ....[@ H61 h97] "<=5
........653 ....[h @10 H10] "<=20
........654 ....[H h14 @37] "<=5
Key: ..@ = anchor (bar/22 24/22), .H = Hit (bar/23* 13/10), .h = hit (bar/23* 11/8)
Areal alternatives for the hits: .Z = reverse split (bar/23* 13/10), .
B = Both (bar/23* 11/8)
