https://youtu.be/R7WQoPS5dVw?t=1855

I found the above position interesting and was wondering what
Jason Pack was going to do. Of course, Axelisation is no longer
a simple algo in match play. But, if my count is correct, the position Axelises
to RD/T. So it must be an even bigger redouble at the match score.
I'm sure Axel would have correctly doubled here, but Jason missed
his connection.

Paul

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

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

https://youtu.be/R7WQoPS5dVw?t=1855

... which translates to

GNU Backgammon Position ID: 2QEAgGwCAAAAAA
Match ID : AQEgAQAAEAAA
+-1--2--3--4--5--6-------7--8--9-10-11-12-+ O: jason (Cube: 2)
OO | O O O O | | | 0 points
OO | O O | | | On roll
OO | | | |
OO | | | |
O | | | |
| |BAR| |^ 9 point match
X | | | |
XX | | | |
XX | X | | |
XX | X X | | |
XX | X X X | | | 2 points
+24-23-22-21-20-19------18-17-16-15-14-13-+ X: mochy
Pip counts: O 21, X 19

in ASCII.

I found the above position interesting and was wondering what
Jason Pack was going to do. Of course, Axelisation is no longer
a simple algo in match play.

Well, it is, but first you have to come up with the doubling window. At
-9/-7 I would not bother wasting precious seconds (or is it rather
minutes?) with match equity calculation, so let us assume that Jason
somehow knew his doubling point of about 48 % and his cash point of
about 77 %. For an efficient cube you want to double right at the cash
point, but there are certainly market losers, so the closer you get to
the end of the game, the closer you will double right at the doubling
point.

If you use Isight with the distinction between long and short races
(last page of my article) it gives an estimate of 69.6 % winning chances
(the truth is about 5 points fewer). Is this close enough to the cash
point at this stage of the game? I have no idea, but probably would
apply Woolsey's Rule and double (maybe Mochy does not want the
volatility of a 4-cube and drops?).

All this holds only for perfect play on both sides. Now the elephant in
the room in the Match Equity Table. I do not know about Jason, but I
certainly have 100 ELO disadvantage against Mochy, very likely more. If
you change GNU Backgammon to the 100-ELO-difference MET of Jacobs and
Trice, then the doubling point immediately decreases to 40 % and the
cash point to 73 %. This amounts to a very strong double.

If only I knew

1. the ELO difference between me and my opponent and

2. how to calculate the doubling window for it

...

So over the board it certainly boils down to hand-waving, gut feeling
and psychology.

Best regards

Axel

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

On Sunday, July 16, 2023 at 11:50:43 AM UTC+1, Axel Reichert wrote:

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

https://youtu.be/R7WQoPS5dVw?t=1855

... which translates to

GNU Backgammon Position ID: 2QEAgGwCAAAAAA
Match ID : AQEgAQAAEAAA
+-1--2--3--4--5--6-------7--8--9-10-11-12-+ O: jason (Cube: 2)
OO | O O O O | | | 0 points
OO | O O | | | On roll
OO | | | |
OO | | | |
O | | | |
| |BAR| |^ 9 point match
X | | | |
XX | | | |
XX | X | | |
XX | X X | | |
XX | X X X | | | 2 points
+24-23-22-21-20-19------18-17-16-15-14-13-+ X: mochy
Pip counts: O 21, X 19

in ASCII.

I found the above position interesting and was wondering what
Jason Pack was going to do. Of course, Axelisation is no longer
a simple algo in match play.

Well, it is, but first you have to come up with the doubling window. At -9/-7 I would not bother wasting precious seconds (or is it rather
minutes?) with match equity calculation, so let us assume that Jason
somehow knew his doubling point of about 48 % and his cash point of
about 77 %. For an efficient cube you want to double right at the cash point, but there are certainly market losers, so the closer you get to
the end of the game, the closer you will double right at the doubling
point.

If you use Isight with the distinction between long and short races
(last page of my article) it gives an estimate of 69.6 % winning chances (the truth is about 5 points fewer). Is this close enough to the cash
point at this stage of the game? I have no idea, but probably would
apply Woolsey's Rule and double (maybe Mochy does not want the
volatility of a 4-cube and drops?).

All this holds only for perfect play on both sides. Now the elephant in
the room in the Match Equity Table. I do not know about Jason, but I certainly have 100 ELO disadvantage against Mochy, very likely more. If
you change GNU Backgammon to the 100-ELO-difference MET of Jacobs and
Trice, then the doubling point immediately decreases to 40 % and the
cash point to 73 %. This amounts to a very strong double.

If only I knew

1. the ELO difference between me and my opponent and

2. how to calculate the doubling window for it

...

So over the board it certainly boils down to hand-waving, gut feeling
and psychology.

Best regards

Axel

Later in that same game, Jason (still with the cube) was in a last-roll position with 26/36
outs (and therefore 10/36 losers). For money this is (of course) an 88.9% take. However,
the match score makes the take much closer here -- it's a 97.1% take.
Mochy took and he probably knew the equity. If Mochy was considering rating differentials, then there might have been a case for dropping.
Basically, I think the consensus is that when Mochy and Jason are partying on, they're both world-class (even though Mochy's stronger of course) so they may as well attempt optimal play rather than making their own play worse.

Paul

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