Thanks to both Don and David for their comments.
To add one idea: voting theory per se attempts to balance additional criteria,
such as prevention of strategic voting, which don't concern us in designing a
game-playing program, since the multiple agents of a Go "Hydra" ( apt choice,
David! ) aren't motivated to subvert the vote; if ever the result of a vote
were to lead to redistribution of resources from one agent to another, that
would become an important design consideration, of course.
The wikipedia article on Borda Voting suggests different strategies for
truncated tickets. It seems unreasonable to require agents to order all 361
possible points on the board from the top to the least. One method is to assign
a minimum value, such as 1 or 0, for anything not specifically voted for.
Another would be for each agent to be assigned a certain number of votes, and
any votes which are not specifically allocated to move candidates would be
evenly distributed amongst the remainder.
I think it would be good for a tactical analysis program to be able to report
"move A is huge, moves B, C, D, and E are worthless, and I don't particularly
care about the rest of the board" -- possibly it would be useful to use Borda
voting to aggregate the results of agents for each viable ( or not-so-viable)
group on the board, along with guesses as to the win probability and shape
factors and other indicators.
( It occurs to me that I recently tweaked a yahoo config switch to quote
replied-to messages. I shall tweak it back and see if this bizarre
one-word-per-line misfeature of yahoo mail goes away. )
Terry McIntyre <[EMAIL PROTECTED]>
____________________________________________________________________________________
Be a better friend, newshound, and
know-it-all with Yahoo! Mobile. Try it now.
http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ
_______________________________________________
computer-go mailing list
[email protected]
http://www.computer-go.org/mailman/listinfo/computer-go/
