With my massive pip count lead, my plan is simply to scoot past XG,
without caring at all about maintaining my board.
Accordingly, I played XG's top play -- 6/1 4/1.
I thought it was essential to maintain the 8 point, and assumed that
breaking the 8 point was a major whopper, and was proud of myself
for realising this fact.
I was therefore quite surprised that it apparently costs less than 0.03 to
lose the 8 point with 8/5 8/3. That looks terrible to me.

Why is losing the 8 point, obviously making it hard to clear, not
such a terrible blunder??

I don't get it.

Paul

XGID=-BB-CBBcB-B--a---abbbbb---:1:-1:1:53:18:26:3:0:10
X:Daniel O:eXtremeGammon

Score is X:18 O:26. Unlimited Game, Jacoby Beaver
+13-14-15-16-17-18------19-20-21-22-23-24-+
| O O O | | O O O O | +---+
| O | | O O O O | | 2 |
| | | | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| O | | X |
| X X O | | X X X X X |
| X X O | | X X X X X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 76 O: 124 X-O: 18-26
Cube: 2, O own cube
X to play 53

1. 4-ply 6/1 4/1 eq:+0.793
Player: 89.90% (G:3.36% B:0.00%)
Opponent: 10.10% (G:0.52% B:0.01%)

2. 4-ply 6/1 5/2 eq:+0.782 (-0.011)
Player: 89.51% (G:3.12% B:0.00%)
Opponent: 10.49% (G:0.52% B:0.02%)

3. 4-ply 6/3 6/1 eq:+0.779 (-0.015)
Player: 89.33% (G:3.13% B:0.00%)
Opponent: 10.67% (G:0.48% B:0.01%)

4. 4-ply 8/5 8/3 eq:+0.767 (-0.027)
Player: 87.98% (G:4.54% B:0.00%)
Opponent: 12.02% (G:0.35% B:0.01%)

5. 1-ply 8/3 6/3 eq:+0.291 (-0.503)
Player: 68.21% (G:5.69% B:0.05%)
Opponent: 31.79% (G:1.89% B:0.01%)


eXtreme Gammon Version: 2.10

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 8/30/2021 7:59 AM, [email protected] wrote:

I was therefore quite surprised that it apparently costs less than 0.03 to lose the 8 point with 8/5 8/3. That looks terrible to me.

A full rollout yields a significantly larger equity difference.

XGID=-BB-CBBcB-B--a---abbbbb---:1:-1:1:53:0:0:3:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game, Jacoby Beaver
+13-14-15-16-17-18------19-20-21-22-23-24-+
| O O O | | O O O O | +---+
| O | | O O O O | | 2 |
| | | | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| O | | X |
| X X O | | X X X X X |
| X X O | | X X X X X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 76 O: 124 X-O: 0-0
Cube: 2, O own cube
X to play 53

1. Rollout¹ 6/1 4/1 eq:+0.825
Player: 91.06% (G:3.23% B:0.00%)
Opponent: 8.94% (G:0.91% B:0.03%)
Confidence: ±0.004 (+0.821..+0.829) - [61.3%]

2. Rollout¹ 6/1 5/2 eq:+0.824 (-0.001)
Player: 91.21% (G:2.76% B:0.00%)
Opponent: 8.79% (G:0.96% B:0.04%)
Confidence: ±0.004 (+0.820..+0.828) - [38.7%]

3. Rollout¹ 6/3 6/1 eq:+0.785 (-0.039)
Player: 89.67% (G:2.22% B:0.00%)
Opponent: 10.33% (G:0.79% B:0.02%)
Confidence: ±0.004 (+0.781..+0.789) - [0.0%]

4. Rollout¹ 8/5 8/3 eq:+0.747 (-0.078)
Player: 87.23% (G:3.49% B:0.01%)
Opponent: 12.77% (G:0.76% B:0.02%)
Confidence: ±0.002 (+0.744..+0.749) - [0.0%]

¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller

eXtreme Gammon Version: 2.19.207.pre-release

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Monday, August 30, 2021 at 1:37:00 PM UTC+1, Tim Chow wrote:

On 8/30/2021 7:59 AM, [email protected] wrote:

I was therefore quite surprised that it apparently costs less than 0.03 to lose the 8 point with 8/5 8/3. That looks terrible to me.

A full rollout yields a significantly larger equity difference.

XGID=-BB-CBBcB-B--a---abbbbb---:1:-1:1:53:0:0:3:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game, Jacoby Beaver +13-14-15-16-17-18------19-20-21-22-23-24-+
| O O O | | O O O O | +---+
| O | | O O O O | | 2 |
| | | | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| O | | X |
| X X O | | X X X X X |
| X X O | | X X X X X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 76 O: 124 X-O: 0-0
Cube: 2, O own cube
X to play 53
1. Rollout¹ 6/1 4/1 eq:+0.825
Player: 91.06% (G:3.23% B:0.00%)
Opponent: 8.94% (G:0.91% B:0.03%)
Confidence: ±0.004 (+0.821..+0.829) - [61.3%]

2. Rollout¹ 6/1 5/2 eq:+0.824 (-0.001)
Player: 91.21% (G:2.76% B:0.00%)
Opponent: 8.79% (G:0.96% B:0.04%)
Confidence: ±0.004 (+0.820..+0.828) - [38.7%]

3. Rollout¹ 6/3 6/1 eq:+0.785 (-0.039)
Player: 89.67% (G:2.22% B:0.00%)
Opponent: 10.33% (G:0.79% B:0.02%)
Confidence: ±0.004 (+0.781..+0.789) - [0.0%]

4. Rollout¹ 8/5 8/3 eq:+0.747 (-0.078)
Player: 87.23% (G:3.49% B:0.01%)
Opponent: 12.77% (G:0.76% B:0.02%)
Confidence: ±0.002 (+0.744..+0.749) - [0.0%]

¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller

eXtreme Gammon Version: 2.19.207.pre-release

---
Tim Chow

Thanks, Tim

Unfortunately, your post merely displaces my disillusionment -- rather than eliminating it.
Now I'm disillusioned by the fact that I can't rely on XG's analysis stats.

Paul

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 8/30/2021 9:31 AM, [email protected] wrote:

Unfortunately, your post merely displaces my disillusionment -- rather than eliminating it.
Now I'm disillusioned by the fact that I can't rely on XG's analysis stats.

Next you just need to learn that you can't rely on XG rollouts either.

---
Tim Chow

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)