| |
BGonline.org Forums
Problem: Cube action?
Posted By: Chuck Bower In Response To: Problem: Cube action? (Daniel Murphy)
Date: Sunday, 16 September 2012, at 2:50 p.m.
A classic argument for this type of position, using Janowski (hopefully correctly, since he's reading+posting on the forum these days :) is that 5::4 underdog is a beaver when there is standard cube ownership equity for the underdog. How far off is this from a pure, ideal 5::4 position?
1) If Blue misses, should White make a roll for the gammon? Lowest pipcount roll that misses is 4-3. I doubt Blue stays around -- that only prevents White from bearing off two in a few cases and allows some joker small doublets. Answer: gammons appear irrelevant.
2) If Blue hits, how often can White still squeeze out a win? Cubeless it's about 5% (of the 55.5% hits), but this is cubeful. In the ideal Keeler-Spencer (and shortly later Zadeh-Kobliska) case, White needs to go from 5% to 80% before Blue needs to push that 5% to zero. This (ideally) happens 5/80 = 1/16 or 6% of the 55.5% = 3.3%.
So even owning the cube it looks like White is a 52::48 underdog.
The reason the ideal Janowski prediction fails (if I did the analysis correctly) is because White gets nowhere near standard recube vig. When Blue fans it's a monster pass and when White hits Blue seldom gets a cube return at all, let alone an efficient one. I.e. this isn't an ideal Janowski (beaver) situation.
OK, it's not a beaver. Is it a double? Whether or not White has been asleep at the wheel in not doubling before now, presumably he'll know to cash next turn. Therefore this is a last roll loss if Blue misses and a last roll win if Blue doesn't double and then hits. If Blue has the advantage and White has a clue about handling the cube going forward then there is no (net) value in waiting.
D+T
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.