I have just read (not particularly thoroughly, admittedly)
Axel's Isight paper. It's simply than I thought, and I look forward
to applying it in practice.

However, the method is (occasionally) not well-defined in my opinion.
But the same lack of definedness is apparent in all methods that I know of. Suppose that a player A (for Alice) has all of her checkers in a single acepoint stack.
Let n be the height of the stack.
Mathematically, this is exactly the same position as if n was replaced by
m = 2 * ceiling(n/2). Yet m and n might well be different.
So which do we use? m or n. Clearly the answer can sometimes matter.
Applying the method literally, the current method is just to use m or n according to how the position is presented on the board. But that
clearly makes no sense whatsoever.
Without further clarification, my gut instincts would be to use m rather than
n since those tend the even stacks tend to be the most common mental
picture when players are creating the analysis. But I'm really not sure.
Any thoughts? Thank you.

Paul

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)