Holding cube while up 6-0 in match to 9. Roll 4-3. Easy decision, right?

[fullerene]

White - Pips 183. Match Score 0/9

Black - Pips 60. Match Score 6/9
Black to Play 4-3
Created with www.BGdiagram.com

[_Z_]

Over the board, I'm sure I would just bear the checker off.

One thing that could be going on though is if you are going to leave a shot, you'd rather leave it early while White still has a lot of checkers up on the bar. So if 7/4 6/2 is only a hair more likely to be hit long-term, it would be right to play that, because the shots tends to come earlier.

[fullerene]

Originally Posted by _Z_

One thing that could be going on though is if you are going to leave a shot, you'd rather leave it early while White still has a lot of checkers up on the bar.

If you bear off on 4-3, what ensuing roll leaves a shot?

[thecantonkid]

In poker, there´s a thing called Fancy Play Sindrome aks FPS. The only play here is to minimize our chances of leaving a shot. With this in mind, we bear a man off.

[jjpregler]

All you have to do is bear off safely. 7/off leaves no shots on your next roll, while leaving your spares on the highest points. Can't really do much better.

[fullerene]

Yeah, bearing off appears obvious, doesn't it? That's what I thought. There is zero risk on the ensuing roll. However, 2-ply yields a 130-error for 7/off (it's "bad"), whereas the rolls with greater equity present some bad combos next time around. I don't understand.

[Eric]

Originally Posted by fullerene

Yeah, bearing off appears obvious, doesn't it? That's what I thought. There is zero risk on the ensuing roll. However, 2-ply yields a 130-error for 7/off (it's "bad"), whereas the rolls with greater equity present some bad combos next time around. I don't understand.

When I put the position to GNUBG, the cubeful 2 ply result was clearly 7/off. I'm not sure why your result would be so different. Did you set the cube correctly? The opponent would have to drop pretty much any recube; whereas they might be able to take an occasional cube from center. That's the only thing I could think of.

[jjpregler]

Originally Posted by fullerene

Yeah, bearing off appears obvious, doesn't it? That's what I thought. There is zero risk on the ensuing roll. However, 2-ply yields a 130-error for 7/off (it's "bad"), whereas the rolls with greater equity present some bad combos next time around. I don't understand.

XG on ++ gives 7/off as the best play. (++ on XG is a mini-rollout) The issue is your program at 2 ply is just wrong. In fact XG gives 7/3 6/3 as 0.61 error.

[fullerene]

Thanks, guys. This is odd. I set the board just as it was before, and 7/off is now listed as the best play. I don't have a clue why GNUBG yielded the equities it did before. Because the analysis confused me so much, I checked, re-checked, and triple-checked to make sure I wasn't missing anything. Nothing on my end explains the error. I guess I'll just chalk it up to a computer glitch.