First of all, things to note about the position. It arose from someone misplaying a 2nd roll. 54S/63? should be played by hitting off the 5pt and since you know this, your opponent will never have to make the decision we're faced with here right?
Anyway, aside from knowing the answer is making the 5pt by a veritable country mile, how do we get there? After an opening 63R if we were to have a 31 to play, what would we play?
If we don't hit and passively make the 5pt here we've essentially allowed our opponent's checker to escape. He'll safety his checker and maintain his racing lead, currently 5 pips + on roll. If we do hit I think it's clear that even though a builder on our 9pt is more valuable than a builder on our 10pt, using the ace to split the back checkers takes precedence over that minor improvement. By hitting we take the initiative, take away half our opponent's roll, bring a builder into our outfield, and split our back checkers giving us more control over our opponent's outfield and more anchor making numbers. Hitting and splitting is clearly correct.
| gnubg | 158 |
|---|---|
| |
| user | 167 |
Position ID: 4HPwCSDgc/ABMA Match ID: cIkFAAAAAAAA | |
| # | Ply | Move | Equity | |||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | R | 24/23 13/10* | +0.0612 | |||||||||||||||||||
| ||||||||||||||||||||||
| Full cubeful rollout with var.redn. | ||||||||||||||||||||||
| 5184 games, Mersenne Twister dice gen. with seed 1016786884 and quasi-random dice | ||||||||||||||||||||||
| Play: 2-ply cubeful prune [world class] | ||||||||||||||||||||||
| keep the first 0 0-ply moves and up to 12 more moves within equity 0.16 | ||||||||||||||||||||||
| Skip pruning for 1-ply moves. | ||||||||||||||||||||||
| Cube: 2-ply cubeful prune [world class] | ||||||||||||||||||||||
| 2 | R | 13/10*/9 | +0.0465 ( -0.0147) | |||||||||||||||||||
| ||||||||||||||||||||||
| Full cubeful rollout with var.redn. | ||||||||||||||||||||||
| 5184 games, Mersenne Twister dice gen. with seed 1016786884 and quasi-random dice | ||||||||||||||||||||||
| Play: 2-ply cubeful prune [world class] | ||||||||||||||||||||||
| keep the first 0 0-ply moves and up to 12 more moves within equity 0.16 | ||||||||||||||||||||||
| Skip pruning for 1-ply moves. | ||||||||||||||||||||||
| Cube: 2-ply cubeful prune [world class] | 3 | R | 8/5 6/5 | +0.0186 ( -0.0426) |
| Full cubeful rollout with var.redn. | 5184 games, Mersenne Twister dice gen. with seed 1016786884 and quasi-random dice | Play: 2-ply cubeful prune [world class] | keep the first 0 0-ply moves and up to 12 more moves within equity 0.16 | Skip pruning for 1-ply moves. | Cube: 2-ply cubeful prune [world class] | |||||||||||
Let's take it back a notch and see if what we expect to happen happens after an opening 62R/31?
Well, since there's no hit & split available the question comes down to is hitting better than making the 5pt? It was in the last rollout so why the hell wouldn't it be here!?
| gnubg | 159 |
|---|---|
| |
| user | 167 |
Position ID: 4HPwESDgc/ABMA Match ID: cIkMAAAAAAAA | |
| # | Ply | Move | Equity | |||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | R | 13/9* | +0.0408 | |||||||||||||||||||
| ||||||||||||||||||||||
| Full cubeful rollout with var.redn. | ||||||||||||||||||||||
| 2592 games, Mersenne Twister dice gen. with seed 1021167104 and quasi-random dice | ||||||||||||||||||||||
| Play: 2-ply cubeful prune [world class] | ||||||||||||||||||||||
| keep the first 0 0-ply moves and up to 12 more moves within equity 0.16 | ||||||||||||||||||||||
| Skip pruning for 1-ply moves. | ||||||||||||||||||||||
| Cube: 2-ply cubeful prune [world class] | ||||||||||||||||||||||
| • | 2 | R | 8/5 6/5 | +0.0230 ( -0.0178) | ||||||||||||||||||
| ||||||||||||||||||||||
| Full cubeful rollout with var.redn. | ||||||||||||||||||||||
| 2592 games, Mersenne Twister dice gen. with seed 1021167104 and quasi-random dice | ||||||||||||||||||||||
| Play: 2-ply cubeful prune [world class] | ||||||||||||||||||||||
| keep the first 0 0-ply moves and up to 12 more moves within equity 0.16 | ||||||||||||||||||||||
| Skip pruning for 1-ply moves. | ||||||||||||||||||||||
| Cube: 2-ply cubeful prune [world class] | ||||||||||||||||||||||
Now to the position that Neil posted, that's on the cover of Ballard/Weaver's new book.
Let's try to note everything that has changed from our 2nd roll decision to our 3rd roll decision.
- Most importantly I believe, we have another spare on our 8pt. Now making the 5pt unstacks two points exactly to the point they'd like to make
- Our back checkers are already split. We have our opponent's golden anchor slotted. With our checkers split our opponent won't be able to safety his outfield blot so easily. We have complete board coverage and by making the 5pt the stronger board. If he is unable to safety that checker or uses his roll in another way, now when & if we hit we'll have a better board. The weenie play of anchoring up is worthless. When you're under no immediate pressure by your opponent it's best to work on offense instead of defense
- We're no longer behind in the race, we're basically even. I believe this to have little effect on the decision but since we're rattling off similarities/differences...
I'm too lazy to get up & double check the book but I believe anything other than making the 5pt is a blunder. 8/5 6/5
| gnubg | 158 |
|---|---|
| |
| user | 158 |
Position ID: 4HPwCSDg8+ABIQ Match ID: cIkMAAAAAAAA | |
