Phil Simborg: I must admit, I missed doubling here, and it clearly is a double for Red. I wasn't so sure it was a double, down 33 pips, so I asked John O'Hagan if he would have doubled this. Here is his response:
Using Woolsey's rule, we first ask if we're sure Blue has a take. ATS Blue's takepoint is 30% plus half of Red's net gammons and Red's 4-cube takepoint will be 25%. This isn't a very gammonish position but Red certainly wins more than Blue. Extreme Gammon's estimate of 17% net gammons for Red looks about right and this raises the takepoint to the 33 percent range.
Does Blue have over 30% cubeless chances in this game? I would certainly think so. He's + 33 in the race, has a bigger board, and Red's an underdog to hit on the mid, make the 7-point, or point on the 2-point. Looks like a pretty clear take for Blue.
Does Red have enough threats to double? He's got quite a few. 44 hits and makes the 7-point, 22 hits or makes the 7-point, while 62 and 53 hit on the midpoint. These 6 numbers are mostly market losers but I don't think they, by themselves, are quite enough to justify a double. Does Red have any other big market losing sequences? Yes I think he does.
Six more numbers point on the 2-point (66,63,33,32) and 43 makes the 7-point. 21 and 11 are good shots as well and, depending on Blue's next roll, might lose the market as well. How is Red doing if he doesn't roll one of these good numbers? He will mostly bring in another builder or maybe hit loose on the 2-point with reasonable chances of a closeout or a prime against Blue's laggard plus a decent 5-point holding game if Blue does squirm out.
My general rule for doubling at normal match scores like this is that doubling is usually correct if 1) You can find 9/36 market losers and 2) You're still in fairly good shape in the other 27/36. I think this position meets both of these conditions and that White should therefore double.