
BGonline.org Forums
Here is the simple answer
Posted By: phil simborg
Date: Monday, 6 October 2014, at 11:16 p.m.
In Response To: Will the real takepoint please stand up? (bill calton)
The take point I use of 22 percent is the "effective live cube take point" which is derived by taking 68 percent of the difference between the live and dead cube take points listed in XG and subtracting that number from the dead cube take point.
This number is very accurate in a race of more than about 60 pips. It allows for reasonable recube vig for a race.
As you know, when the cube is turned, the Gammon value is 1.0 at this score. This is where things get dicey. Let's say you have a position where you can win the game 45 percent of the time but you get gammoned 30 percent of the time.
If you add 30 to 22 you get a gammonadjusted TP of 52, so it would be a big drop.
However, this is a false number, because if you drop the cube your MWC is 42.85, so you would be better off taking the cube and immediately recubing and win the game and match 45 percent of the time (obviously when you recube the gammons no longer matter).
So you can use a very complicated formula to calculate your true recube vig (I have it explained in a twopage paper from John O'Hagan that is brilliant but I need an interpreter to fully understand it).
Or you can use a very new formula that I have worked out with the help of my student, Paul Donnellan from England. At this score, after you compute the gammon value, you reduce the gammon price by 25 percent (because of the recube vig) and add that to the 22 percent effective live cube take point, and you would get a closer approximation of the gammonadjusted take point.
Again, you don't have to do any of this if you estimate that your winning chances are higher than your MWC if you were to drop. By the way, at this score you could actually take at a few points under 42.85...say, around 40 percent, because you have some cube vig if you wait to redouble at a reasonably efficient time.
Now, keep in mind that just because you can take and recube immediately to wipe out gammons, that doesn't mean you should redouble immediately. As long as you are not losing your market, it's best to wait until you have an efficient cube...you might get a wrong take or pass, and you might never recube and just lose 2 points and still be in the game (with 33 percent winning chances).
Now, relative to my 22 percent number, after years of discussing this with Xavier, he has agreed that the next version of XG will have a third column between the dead and live cube take points that list this effective number.
I have tested my effective numbers in hundreds, maybe thousands of positions when you count the positions where my students and teaching partners (Stick, John, Perry and David) have also checked them, and they work very well for most positions. You need to exercise judgment as to which kinds of positions have more or less cube vig (racing positions tend to have the most cube vig and many highly volatile positions and end game cubes have little or no cube vig) and adjust the numbers, but they work well, and unless you are a John O'Hagan or Art Benjamin and can do all the crazy math over the board to factor in cube vig, I strongly recommend you go with this approach.

BGonline.org Forums is maintained by Stick with WebBBS 5.12.