The three is forced. Where to play the four?
Position ID: sZ2FASKG14wDQA
Match ID: cAmuAAAACAAA
Spoiler
The runner is way utg, and advancing it to the bar point leaves a triple shot. We're on the ropes here, so the best option seems to be an action play that takes half a roll away from our opponent. Hit loose and pray.
Spoiler
Why 7/3?
Bonus points for explaining how either it or 8/4* comports with the cube that's probably coming our way momentarily.
Spoiler
I've changed my mind. I think we should try running and recycle if we have to, which does seem likely, whatever we do.
Spoiler
i like 8/4* too, then we must decide if we want to accept the cube? ;)
Spoiler
IMHO the idea here is to make a play that'll get us doubled so we can quietly drop. If we give him his whole move he's liable to play for a gammon. He may do so regardless of our move.
I think we hit him with the 4 and hope the idea of a half-move persuades him to turn the cube.
Bob
Quote from: NIHILIST on October 23, 2010, 10:14:15 PM
Spoiler
IMHO the idea here is to make a play that'll get us doubled so we can quietly drop. If we give him his whole move he's liable to play for a gammon. He may do so regardless of our move.
I think we hit him with the 4 and hope the idea of a half-move persuades him to turn the cube.
Bob
Spoiler
This is a brilliant manifestation of 'strategic vision'. Without getting engulfed in number-crunching fantasies, you offered a simple, yet efficient and ellegant escape route for this miserable herd at this historical moment.
Thou shall be our Shepherd Elect in our plight against this bald wolf. May the Allah bless you, Leviathan!
PL
I don't know how brilliant it was, since it appears he didn't fall for it.
Bob
expurgated bot results
Spoiler
I have gnu 0-ply rollout, but it's too soon to present it in it's entirety.
The Herd chose the best move here. 7/3 is second best and only a .02 error - basically within the margin of error. The other plays are double-whoppers (.2 errors)
XG rollout:
Spoiler
Both plays lead to a big double but also an easy take. Very similar results for both plays (to my surprise), except that the hit is just a little bit better in every aspect.
XGID=aA--BBC-BA--cB-b-bad-A-bA-:0:0:-1:43:0:1:0:5:10
1. Rollout¹ Bar/22 8/4* eq:-0,789
Player : 38,47% (G:15,10% B:3,04%)
Opponent: 61,53% (G:33,51% B:2,89%)
Confidence: +/- 0,013 (-0,802<E<-0,776)
Duration: 9 hours 52 minutes
2. Rollout¹ Bar/22 7/3 eq:-0,811 (-0,023)
Player : 38,14% (G:14,28% B:2,82%)
Opponent: 61,86% (G:34,50% B:0,97%)
Confidence: +/- 0,012 (-0,823<E<-0,799)
Duration: 7 hours 48 minutes
¹ 10368 Games rolled with Variance Reduction.
Moves and cube decisions: 3 ply
Search interval: Huge
eXtreme Gammon Version: 1.21, MET: Rockwell-Kazaross