Poll
Question:
6-3
Option 1: 20/ 11
votes: 8
Option 2: 20/14 9/6
votes: 3
Option 3: 20/17 9/3
votes: 0
Option 4: 9/3 4/1
votes: 0
Option 5: 9/3 6/3
votes: 0
Run or stay and fight?
Spoiler
I'm inclined run with 20-11. Two crossovers, leaving us with just two to bring in to Schig's four and him with potential gapage. Might be wrong though; I could be persuaded to change my mind.
Spoiler
Time to run; close race, close crossovers. To stay is to start wasting pips and likely suffer the indignation of losing by pick - n - pass...........like being nibbled to death by gerbils.
20-14, 9-6.
Bob
Spoiler
It's clearly right to go. Now our blot is just a target for red. Only (!) 6-2 leaves a shot next turn and there are all those ( count them yourself) numbers that pick and pass or close us out instantly. Also we lead in the race after the roll. Either running play may be correct. I have played 20/11, but 20/14, 9/6 might be better, I haven't a clue about that.
Spoiler
Transitioning to a running game seems clear - we're ahead by a tempo, and we have the cube so we're actually favored to win. Staying behind? i can't think of anything positive.
So 20/14 is clear. Where to play the 3? I like moving a checker into the home board. For one thing, any 9 will bring the back man home, but with 20/11 5-4 still leaves a man in the outfield.
That said, I don't think there's a big difference between 20/11 and 20/14 9/6. Either is fine.
Spoiler
I'm just a crossover purist, I'm sure they're both close.
Bob
Quote from: NIHILIST on December 19, 2009, 10:32:23 PM
...I'm sure they're both close.
Yeah, so am I, and although I'm with you on the choice of move, it's not worth prolonging the vote over a small equity margin. So I'm going to move it along. The real choice is whether to stay or go and we're unanimous on that.
I went for 20/11. My reason for that instead of 9/6 is because we have the cubed and the rhthem going for us now. Since we have the cubed we are a favorite to win this game. We are not trying to get of gammon here. If we move 9/6 we dont have any flexibilty on how we move the next roll. With 20/11 we have more options next turn. It is pretty much the same though and since no one is voting for the stay back and fight play there i no need to denounce it. It has to be a whopper of a blunder to stay. I only see 6-2 as the only blot number.. 5-4,44,22,55 close us out and 41,42,43,51, 52, 53, and are pick n pass. 11 is a pick and slide.
plus if we run we are 3 pips better than %50 in the race so we are about a 56% favorite plus we have the cube. A redouble will put us out perfectly. I think schig will have to be 32% to take.. eventhough hell have an auto rere to 8. S after running we only have to advance like 12-13% more to win the game. This is clearly more of a next then immediately meets the eye. (i mean next between "fight and run")
Agree that it's a blunder to stay.
But I think our advantage is more than just having the cube - we are ahead in the race by 7 pips, which is basically the same as being on roll and even. Worth a tempo. Of course, having access to the cube is worth something too, perhaps as much or more as the tempo.
Like Bob, I'm a crossover purist and will move a checker into the homeboard before making any other transition. My estimate is that 9/6 is a small equity gain over 20/11, but small enough that it's not worth quibbling about. Still, I'm surprised that the vote was this lopsided.
I'm going to defer my cube analysis until after he's moved; suffice it to say that we'll have a cube vote no matter how he plays this roll. I don't think that gives away much....
Expurgated bot results:
Spoiler
I've snipped the absolute equities and presented only the relative equities so as not to give away data that might be useful to a cube decision.
As expected the top two plays are close - a statistical tie.
1. Rollout 20/14 9/6 Eq.:
2. Rollout 20/11 Eq.: ( -0.006)
3. Rollout 20/14 6/3 Eq.: ( -0.090)
Full cubeful rollout with var.redn.
1296 games, Mersenne Twister dice gen. with seed 840571151 and quasi-random dice
Play: 0-ply cubeful prune [expert]
Cube: 0-ply cubeful prune [expert]