| |
BGonline.org Forums
SNOWIE ROLLOUTS
Posted By: Matt Ryder In Response To: SNOWIE ROLLOUTS (Timothy Chow)
Date: Thursday, 7 January 2010, at 6:14 a.m.
In particular, I believe that "live-cube" in the context of evaluations means that the cube action is assumed to be perfectly efficient (i.e., that we use the version of Janowski in which the cube is assumed to be always turned at the take/drop borderline), whereas in the context of rollouts it means that the position of the cube is being tracked as the game evolves.
I don't agree with your statement about "live-cube" efficiency. In both a static evaluation and a rollout, a changing 'cube efficiency' variable X is substituted depending on the position. See my recent post. And in both cases a Janowski formula is applied iteratively n-plies forward to establish a cubeful equity from the cubeless figures at each point.
Anyway, the point I was making is that you shouldn't expect to be able to compute anything other than the cubeless equity from the 6 probabilities.
This is where we part company. As I understand it you can most certainly derive a cubeful equity from the 6 cubeless probabilities. The whole point of Janowski's formula is to provide a mathematical transformation that will achieve exactly this. As I understand it, the size of the cube is normalized in the display of the cubeful equity so that a win or loss equates to 1/-1. So I think cubeful equity indicates the impact of the cube on the outcome of the game, but says nothing directly about the size of the cube. (I'm only considering money games here).
For your interest, here's Janowski's original paper: Take-Points in Money Games. As a mathematician, you should make substantially better sense of it than I did.
Of course, I've only just started investigating this stuff and am finding it difficult to see the forest for the trees. As such, perhaps I'm barking up the wrong one.
Matt R.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.