I wrongly dropped the below, using Isight, but the drop appears wrong.
However, if I was operating pre-Axelishly, using my own knowledge
and experience, I would have dropped anyway.
So maybe my claim for compensation should be denied. Before I used
Isight, my thinking would be this:
"A pure 4 roll vs 4 roll position is a slim take. It's close to what we have
here, but is it better or worse? I have some ugly gaps and I'm a pip down.
I'll pass." I wouldn't know the EPC, but even if I did, it wouldn't
seem like take territory.

BTW, I do make fun non-financial bets sometimes. They work like this:
The winner of the bet picks a pdf or URL available online
that is interesting to the winner, and the loser of the bet is obligated to
read the material.
Thus, if you win enough bets, you force other people to read up on stuff
that you're interested in, which I think is quite rewarding.

Paul

XGID=--BD-A---------------bcaa-:4:-1:-1:00:0:0:3:0:10
X:eXtremeGammon O:Daniel

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 |
| | | |
| |BAR| |
| | | |
| | | |
| | | X | +---+
| | | X X | | 16|
| | | X X X X | +---+
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 20 O: 21 X-O: 0-0
Cube: 16, X own cube
X on roll, cube action

Analyzed in Rollout
No redouble
Player Winning Chances: 75.33% (G:0.00% B:0.00%)
Opponent Winning Chances: 24.67% (G:0.00% B:0.00%)
Redouble/Take
Player Winning Chances: 75.33% (G:0.00% B:0.00%)
Opponent Winning Chances: 24.67% (G:0.00% B:0.00%)

Cubeless Equities: No Double=+0.507, Double=+1.013

Cubeful Equities:
No redouble: +0.793 (-0.140)
Redouble/Take: +0.933
Redouble/Pass: +1.000 (+0.067)

Best Cube action: Redouble / Take

Rollout:
1296 Games rolled with Variance Reduction.
Moves: 3-ply, cube decisions: XG Roller
Confidence No Double: ± 0.002 (+0.791..+0.795)
Confidence Double: ± 0.003 (+0.930..+0.935)

Double Decision confidence: 100.0%
Take Decision confidence: 100.0%

Duration: 2.3 seconds

eXtreme Gammon Version: 2.10

"peps...@gmail.com" <pepstein5@gmail.com> writes:

> I wouldn't know the EPC, but even if I did, it wouldn't
> seem like take territory.

[...]

> XGID=--BD-A---------------bcaa-:4:-1:-1:00:0:0:3:0:10
> X:eXtremeGammon O:Daniel
>
> 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 |
> | | | |
> | |BAR| |
> | | | |
> | | | |
> | | | X | +---+
> | | | X X | | 16|
> | | | X X X X | +---+
> +12-11-10--9--8--7-------6--5--4--3--2--1-+
> Pip count X: 20 O: 21 X-O: 0-0
> Cube: 16, X own cube
> X on roll, cube action

Just to explore what you might get with the (abandonned, because much
worse than the real Isight method) EPC approximation from my article
(page 20):

EPC(X) = 20 + 5 + 2*1 + 1*1 + 1*3 - 1 = 30 (exact 30.86)
EPC(O) = 21 + 5 + 2*0 + 1*2 + 1*4 - 0 = 32 (exact 31.94)

Point of last take is:

30 + 30/8.167 - 3 = 30.67 < 32 => Drop

Even with exact EPCs, the point of last take is:

30.86 + 30.86/8.167 - 3 = 31.67 < 31.94 => Drop

That's how it goes ...

Axel

On Sunday, April 24, 2022 at 4:58:22 PM UTC+1, Axel Reichert wrote:
> "peps...@gmail.com" <peps...@gmail.com> writes:
>
> > I wouldn't know the EPC, but even if I did, it wouldn't
> > seem like take territory.
> [...]
> > XGID=--BD-A---------------bcaa-:4:-1:-1:00:0:0:3:0:10
> > X:eXtremeGammon O:Daniel
> >
> > 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 |
> > | | | |
> > | |BAR| |
> > | | | |
> > | | | |
> > | | | X | +---+
> > | | | X X | | 16|
> > | | | X X X X | +---+
> > +12-11-10--9--8--7-------6--5--4--3--2--1-+
> > Pip count X: 20 O: 21 X-O: 0-0
> > Cube: 16, X own cube
> > X on roll, cube action
> Just to explore what you might get with the (abandonned, because much
> worse than the real Isight method) EPC approximation from my article
> (page 20):
>
> EPC(X) = 20 + 5 + 2*1 + 1*1 + 1*3 - 1 = 30 (exact 30.86)
> EPC(O) = 21 + 5 + 2*0 + 1*2 + 1*4 - 0 = 32 (exact 31.94)
>
> Point of last take is:
>
> 30 + 30/8.167 - 3 = 30.67 < 32 => Drop
>
> Even with exact EPCs, the point of last take is:
>
> 30.86 + 30.86/8.167 - 3 = 31.67 < 31.94 => Drop
>
> That's how it goes ...
>
> Axel

The only way anyone can understand that this is a take is if they can somehow
realise that the winning probability for the roller is < 78%.
Since it's quite a lot less --- approx 75%, I would expect Stick and his friends to
correctly take.

Paul