[EM] How to fix the flawed "Nash equilibrium" concept for voting-theory purposes

[Jameson Quinn]

My proposal resolves most of those issues, after the votes are cast,
> each ballot has a probability of p to be excluded from the count.
>

That works out to be the same as the poisson proposal, in the limit as voter
number -> infinity and p -> 1. I think that the poisson proposal leads to
cleaner math. For instance, just for the sake of argument, imagine a system
where a margin of 1 counts as a tie and a coin is flipped. In this system,
with your proposal, a single vote might not be enough to increase the
expected value, because it is the only vote of its kind. However, with a
poisson distribution, 1 vote can turn into arbitrarily many (with a very low
probability).
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20100413/72838e7a/attachment-0004.htm>