This seems interesting, but I'm afraid I don't see the difference between Magnetic Stones and the traditional idea of influence. Is there one?
HolIgor: From my exoperience, the way to knowledge is in reformulation of the main principles in your own understandable terms.
Some years ago I saw a derivation for the possible omission number, which is the number of opposing threats to a group's safety that the defender can ignore before needing to respond. Not a very useful feature, perhaps, but interesting. What was very interesting was that there was apparently a good way to estimate that number. How many threats you can ignore does seem to have some relation to influence, doesn't it?
People keep trying, and I think that someday someone will succeed in defining a very good influence function for go.
Charles Hmmm ... I could imagine a function of positions on the side - unsettled groups with some sort of base - which did what Bill suggested on threats, by a combinatorial algorithm. What is more it could be a useful, insightful thing to have, in the way of fine-tuning one's intuition.