                         |
| ||||
| XGID=---aa-EAC--abCd-b--bA-AbA-:0:0:1:21:0:0:3:0:10 | |||||
| Blue to play 21 | |||||
I wasn't sure about which play was best for money and DMP. I guessed that 5P might lose a few too many Gs so made that play at DMP where that wouldn't matter.
It seems I was wrong about the G losses.
Can someone provide an explanation of why 13/11, 8/7 is best at DMP?
I interpret your fuller question as: Why is D (Down, 13/11 8/7) best at DMP and by contrast P (Point, 7/5 6/5) best for Money? Well, you could look at the gammon wins and losses in the details of the rollouts. As you implied, your intuitive estimation was probably off there. But obviously you thought of that and are looking for something more substantial to boost your intuition going forward.
It's a good question. Fortunately, a useful answer is not out of reach.
The only complete answer (to any such question) would need to include a look ahead and tally of all possible game outcomes that can occur from this point forward, unless the asker of the question can be convinced by a less-huge number of evaluations (assuming they can be provided) less far along the tree. Obviously, even if possible this would be a highly impractical approach.
Even looking one roll ahead may not offer the most effective aid. For example, why is it right to hit on the 7pt with 62S-42? We might regale an inexperienced player with what the opponent's numbers do if he doesn't hit and compare them with what they do if he does hit, and he (for that matter, we all) can learn from it, but it is likely to make his head spin.
A comparison to a similar position is usually more helpful. If we point out that although with 62S-42 we hit on the 7pt. with 62S-31 we do NOT hit on the 7pt, and that the reason is because the 5pt is that much more valuable than the 4pt, well... then we've conveyed a bite-sized explanation as well as having provided another useful reference position. Does it "prove" that 62S-42H is better than 62S-42P? Of course not.
Four score positions are very valuable studies for comparing different types of plays at different scores. Here is a complete set of links to R4s (real four-score positions) that have been discovered thus far:
11a.. 11b.. 22a+b.. 33.. 44a.. 44bcdefghijk.. 55.. 66.. 21.. 41.. 43.. 53.. 64.. 65..
The 21, 43 and 65 positions support making the 5pt for Money. This is a strong move for winning gammons while being close to the winningest play, AND with an affordable percentage of gammon losses. In the case of 43, White has ten checkers in the zone but Blue's three back checker blots are far back. In the case of 65, White has only eight in the zone.
The case of the 21 position (click or see diagram at the top of this post) is more subtle. White has ten in the zone but four of them are way back on the 11pt. If the smallest change is made, which would be moving one of those checkers forward to the 9pt, the Money rollout lead of P (Point, 7/5 6/5) over D (13/11* 8/7) reduces from .011 to .004. If that same checker is moved to the 6pt, P becomes .023 worse than D and a .048 error overall -- buttoning up with B (Both, 22/20 8/7) becomes easily best (though D still dominates at DMP). This may give you a better idea of the sensitivity of spare placement when evaluating the risk of an opponent blitz.
When contrasted to other scores, DMP tends to support plays that hit (44a, 55, 21, 43, 64), and/or create an anchor in the opponent's outer board -- most typically the 18pt (33, 55, 66, 65), and/or run (44a, 44bcdefghijk, 55, 43, 53), and or prime (11b, 22a+b, 66, 21). I believe you will find a review of these positions highly instructive, as did I.
Nack