Another noncontact cube action

Timothy Chow

unread,

Jul 21, 2022, 8:57:11 AM7/21/22

to

XGID=---CCCE----Aa---a-ad-ec---:0:0:1:00:0:0:0:0:10

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 |
| | | O O O |
| | | O O O |
| | | O O |
| | | O |
| |BAR| |
| | | X |
| | | X |
| | | X 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: 77 O: 82 X-O: 0-0
Cube: 1
X on roll, cube action

---
Tim Chow

ah....Clem

unread,

Jul 21, 2022, 4:44:29 PM7/21/22

to

Adj pipcount is 78 - 86, POLT is 8, so looks like a close take and a
very clear cube. Won't be surprised if it's a close pass.

--
Ah....Clem
The future is fun, the future is fair.

Timothy Chow

unread,

Jul 23, 2022, 8:47:14 AM7/23/22

to

This time the Keith count says it's a clear ND, which is wrong
but not wildly wrong. I think it's because O's distribution has
some flaws that the Keith count fails to account for. Does Isight
have a different verdict? I have the vague impression that Isight
is designed for positions where all the checkers are already in the
home board, but maybe that impression is wrong.

Analyzed in Rollout
No double
Player Winning Chances: 73.48% (G:0.00% B:0.00%)
Opponent Winning Chances: 26.52% (G:0.00% B:0.00%)
Double/Take
Player Winning Chances: 73.47% (G:0.00% B:0.00%)
Opponent Winning Chances: 26.53% (G:0.00% B:0.00%)

Cubeful Equities:
No double: +0.726 (-0.041)
Double/Take: +0.767
Double/Pass: +1.000 (+0.233)

Best Cube action: Double / Take

Rollout:
1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 4-ply, cube decisions: XG Roller+
Search interval: Huge
Confidence No Double: ± 0.002 (+0.724..+0.728)
Confidence Double: ± 0.003 (+0.764..+0.770)

eXtreme Gammon Version: 2.19.207.pre-release

---
Tim Chow

peps...@gmail.com

unread,

Jul 23, 2022, 7:28:01 PM7/23/22

to

Ok. I'll Axelise this one. But I'm only doing it from my memory
of the algo so results aren't 100% guaranteed, but I've done
a fair few so hopefully I've got the hang of it.

No adjustments for us.
For the opponent, there's +1 for the 5point gap and + 3 for the 3 extra crossovers.
So we get 77 to 86. Add 1/6 to our 77 to get 89 5/6. The difference between 89 5/6 and 86
is 3 5/6. Since this is >=2 the opponent will take. Since this is <=4, we would double even if
we owned the cube.

So the Axelisation works correctly here. In fact, it seems to understand the position very well.
Big take, marginal redouble.

Your impression about Isight is quite egregiously wrong. A full database of non-contact positions
in real games has been consulted, with no preference given to positions where all checkers are in
the home board.
Isight counts crossovers and Keith doesn't, which amply explains why Isight is more positive about
the on-roll player than Keith is.

Paul

peps...@gmail.com

unread,

Jul 24, 2022, 5:53:40 AM7/24/22

to

Could you possibly let me know if this is a redouble?
For some reason, I am unable to set up positions in XG.
I do a lot of rollouts but I can only do these from my games.

Thanks, Paul

Timothy Chow

unread,

Jul 24, 2022, 8:49:48 AM7/24/22

to

On 7/23/2022 7:28 PM, peps...@gmail.com wrote:

> Your impression about Isight is quite egregiously wrong. A full database of non-contact positions
> in real games has been consulted, with no preference given to positions where all checkers are in
> the home board.

I've figured out the source of my confusion. I was confusing Isight
with Joachim Matussek's work.

https://bkgm.com/articles/Matussek/BearoffGWC.pdf

---
Tim Chow

Timothy Chow

unread,

Jul 24, 2022, 9:37:16 AM7/24/22

to

On 7/24/2022 5:53 AM, peps...@gmail.com wrote:
> Could you possibly let me know if this is a redouble?
> For some reason, I am unable to set up positions in XG.
> I do a lot of rollouts but I can only do these from my games.

Possibly a marginal redouble according to the rollout below, but
only just.

XGID=---CCCE----Aa---a-ad-ec---:1:1:1:00:0:0:0:0:10

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 |
| | | O O O |
| | | O O O |
| | | O O |
| | | O |
| |BAR| |
| | | X |
| | | X |

| | | X X X X | +---+
| | | X X X X | | 2 |
| O X | | X X X X | +---+

+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 77 O: 82 X-O: 0-0

Cube: 2, X own cube

X on roll, cube action

Analyzed in Rollout
No redouble
Player Winning Chances: 73.49% (G:0.00% B:0.00%)
Opponent Winning Chances: 26.51% (G:0.00% B:0.00%)
Redouble/Take

Player Winning Chances: 73.47% (G:0.00% B:0.00%)
Opponent Winning Chances: 26.53% (G:0.00% B:0.00%)

Cubeful Equities:

No redouble: +0.763 (-0.004)
Redouble/Take: +0.767
Redouble/Pass: +1.000 (+0.233)

Best Cube action: Redouble / Take

Rollout:
1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 4-ply, cube decisions: XG Roller+
Search interval: Huge

Confidence No Double: ± 0.002 (+0.761..+0.765)

Confidence Double: ± 0.003 (+0.764..+0.770)

eXtreme Gammon Version: 2.19.207.pre-release

---
Tim Chow

Stick Rice

unread,

Aug 4, 2022, 2:11:35 PM8/4/22

to

Keith Count says it's a clear ND? First off, isn't it only by 1 pip? Second off, I've posted it many times that the Keith Count at longer lengths should be adjusted.

~80-100 pips: double when leading by 5
~100-120 pips: double when leading by 6
~120-140 pips: double when leading by 7

Stick