[EM] simulation of Approval, Schulze, and sincere utility

[Kevin Venzke]

I've written a little election generator which generates factions and their
sincere utilities for the candidates.  It then finds the candidate who wins by
highest average utility, Approval, and Schulze.

It assumes no one truncates in Schulze.  For Approval, the voters use "better-
than-expectation" strategy, and there are two options for calculating the
candidates' odds:

1. every voter believes candidates A and B have a 50% chance each.
2. every voter believes each candidate's odds of winning are equal to his odds
of winning by Random Ballot.  That implies that everyone knows everyone else's
sincere favorite.

It occurs to me to run simulations to see whether Schulze or Approval (under
these random circumstances) is better at maximizing utility, and by how much.
I'm not sure what else to do with this, though.