Priming battle

[Timothy Chow]

XGID=-aBBB-BBA--B---b-bbdcB-a--:0:0:1:51:0:0:0:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| O O O | | O O X O |
| O O O | | O O X |
| | | O O |
| | | O |
| | | |
| |BAR| |
| | | |
| | | |
| | | |
| X X | | X X X X |
| X X X | | X X X X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 116 O: 115 X-O: 0-0
Cube: 1
X to play 51

---
Tim Chow

[Timothy Chow]

X is in big trouble if O rolls a 4 regardless, and slotting gives
X the biggest upside if O does not roll a 4. Nevertheless, I was
surprised at how big an error XG thought my play of 11/10 11/6 was.

1. Rollout¹ 11/5 eq:+0.527
Player: 65.18% (G:12.78% B:0.33%)
Opponent: 34.82% (G:7.23% B:0.24%)
Confidence: ±0.008 (+0.519..+0.535) - [100.0%]

2. Rollout¹ 11/6 8/7 eq:+0.257 (-0.270)
Player: 57.96% (G:10.80% B:0.29%)
Opponent: 42.04% (G:6.61% B:0.24%)
Confidence: ±0.011 (+0.246..+0.268) - [0.0%]

3. Rollout¹ 11/10 11/6 eq:+0.193 (-0.334)
Player: 56.39% (G:10.13% B:0.24%)
Opponent: 43.61% (G:6.00% B:0.20%)
Confidence: ±0.010 (+0.183..+0.203) - [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

[Stick Rice]

With your play when the opponent doesn't roll a four you make your five point and a six prime cleanly about 1/4th of the time. So out of the 25 games the opponent doesn't roll a four you are making your five point about 6 games.

By slotting in the games the opponent doesn't roll a four you make your five point and a six prime cleanly about 3/4th of the time. So out of the 25 games the opponent doesn't roll a four you are making your five point about 19 games. This play also leaves you a good six to play which with the other play is otherwise awful, and doesn't duplicate your own fives. So it's no surprise the huge disparity between the plays.

Stick

[Timothy Chow]

On 5/20/2022 2:56 PM, Stick Rice wrote:
> By slotting in the games the opponent doesn't roll a four you make your five point and a six prime cleanly about 3/4th of the time. So out of the 25 games the opponent doesn't roll a four you are making your five point about 19 games. This play also leaves you a good six to play which with the other play is otherwise awful, and doesn't duplicate your own fives. So it's no surprise the huge disparity between the plays.

I wouldn't have been surprised by a .2 disparity, but .3+ surprises me.

---
Tim Chow

[atila...@gmail.com]

21 Mayıs 2022 Cumartesi tarihinde saat 18:05:00 UTC+3 itibarıyla Tim Chow şunları yazdı:

The right play and theme is 11/5. And Stick Rice’s comment about this move is quite satisfying. But failure to grasp the main theme is not the only reason why the error is so great. I think there are 3 more important points.

The fist is O’s blot on the 23 point. The checker duplicates O’s fours. In variant 1, I moved that checker from 23 to 22. And I removed it in variant 2.

Variant 1

XGID=-aBBB-BBA--B---b-bbdcBa---:0:0:1:51:0:0:3:0:10

1. Rollout¹ 11/5 eq:+0,470
Player: 62,77% (G:12,67% B:0,31%)
Opponent: 37,23% (G:7,67% B:0,29%)
Confidence: ±0,008 (+0,462..+0,478) - [100,0%]

2. Rollout¹ 11/6 8/7 eq:+0,271 (-0,199)
Player: 57,30% (G:11,07% B:0,32%)
Opponent: 42,70% (G:6,32% B:0,25%)
Confidence: ±0,012 (+0,258..+0,283) - [0,0%]

3. Rollout¹ 11/10 11/6 eq:+0,216 (-0,254)
Player: 56,14% (G:10,82% B:0,30%)
Opponent: 43,86% (G:6,37% B:0,24%)
Confidence: ±0,012 (+0,204..+0,227) - [0,0%]

¹ 1296 Games rolled with Variance Reduction.

Moves: 3-ply, cube decisions: XG Roller

Variant 2

XGID=-aBBB-BBA--B---b-bbddB----:0:0:1:51:0:0:3:0:10

1. Rollout¹ 11/5 eq:+0,462
Player: 62,87% (G:13,19% B:0,32%)
Opponent: 37,13% (G:6,83% B:0,20%)
Confidence: ±0,008 (+0,453..+0,470) - [100,0%]

2. Rollout¹ 11/6 8/7 eq:+0,308 (-0,154)
Player: 58,48% (G:11,40% B:0,31%)
Opponent: 41,52% (G:6,12% B:0,23%)
Confidence: ±0,013 (+0,295..+0,320) - [0,0%]

3. Rollout¹ 11/10 11/6 eq:+0,224 (-0,237)
Player: 56,40% (G:11,00% B:0,28%)
Opponent: 43,60% (G:6,02% B:0,22%)
Confidence: ±0,012 (+0,212..+0,237) - [0,0%]

Second, John's timing is shaky. The third is the location of the cube. In the third variant, I made all the changes simultaneously. In other words, the cube is now at 2, no blots, and O has 2 checkers on the mid-point to play comfortably.

Variant 3

XGID=-aBBB-BBA--Bb--b-bbccB----:1:-1:1:51:0:0:3:0:10

X:Player 1 O:Player 2

Score is X:0 O:0. Unlimited Game, Jacoby Beaver

Pip count X: 116 O: 133 X-O: 0-0
Cube: 2, O own cube
X to play 51

1. Rollout¹ 11/5 eq:-0,138
Player: 50,27% (G:15,21% B:0,35%)
Opponent: 49,73% (G:8,90% B:0,38%)
Confidence: ±0,010 (-0,148..-0,128) - [100,0%]

2. Rollout¹ 11/6 8/7 eq:-0,199 (-0,062)
Player: 48,59% (G:14,13% B:0,29%)
Opponent: 51,41% (G:8,30% B:0,37%)
Confidence: ±0,011 (-0,210..-0,189) - [0,0%]

3. Rollout¹ 11/10 11/6 eq:-0,237 (-0,100)
Player: 47,09% (G:13,04% B:0,20%)
Opponent: 52,91% (G:7,63% B:0,34%)
Confidence: ±0,010 (-0,247..-0,227) - [0,0%]

[Timothy Chow]

On 5/21/2022 6:04 PM, atila...@gmail.com wrote:
> Second, John's timing is shaky.

Nice variants. But which player are you calling John?

---
Tim Chow

[atila...@gmail.com]

23 Mayıs 2022 Pazartesi tarihinde saat 15:36:36 UTC+3 itibarıyla Tim Chow şunları yazdı:

I don't know John :)) He is O