is Player 2 score: 0 pip: 4 | ||||||||||||||||
Unlimited Game Jacoby Beaver | ||||||||||||||||
pip: 4 score: 0 is Player 1 | ||||||||||||||||
XGID=--B--------------------b--:1:-1:-1:00:0:0:3:0:10 | ||||||||||||||||
on roll, cube action? |
Analyzed in XG Roller++ | No redouble | Redouble/Take |
Player Winning Chances: | 79.94% (G:0.00% B:0.00%) | 79.94% (G:0.00% B:0.00%) |
Opponent Winning Chances: | 20.06% (G:0.00% B:0.00%) | 20.06% (G:0.00% B:0.00%) |
Cubeless Equities | +0.599 | +1.198 |
Cubeful Equities | ||
No redouble: | +0.599 (-0.352) | |
Redouble/Take: | +0.951 | |
Redouble/Pass: | +1.000 (+0.049) | |
Best Cube action: Redouble / Take |
eXtreme Gammon Version: 2.02
We all know this is a double and a take.
But why? Yellow (the weaker side) only wins 20%. Why take? Because of the recube.
OK, that makes some sense. But how do you know? How do you measure that?
This is what I learned in Cleveland. Last roll position.
Win chances = [(# rolls opponent misses) x (# rolls you hit)] / 1296.
You need 25% of 324 to take.
(10) * (26) = 260 implies drop.
What gives? OK, so this is the trick. The recube doubles your equity to the extent you are a favorite.
So, when we get to roll we have 26 rolls to get off, or 8 more than half of 18. We add the 8 rolls into the calculation.
It's as if we get off 34 times (rather than 26) since we win those games with a recube.
So, the corrected calculation is:
(10) * (26 + 8) = 340 or take.
We get double the value on those wins.
Pretty neat. Thanks Dima. dfd