[EM] debian data 2001-2005

[Warren Smith]

http://math.temple.edu/~wds/crv/Debian2003.html
now summarizes it.

All methods agree on all winners in all years 2001-2005 exceot for
2003.

And except for range in 2005 picked Towns as the winner.

So in both 2003 and 2005, range disagreed on the winner with every other method
robla programmed (using robla's standard scheme for converting to range ballots,
anyhow).  This is because the other methods all ignored intensity of preference
information.

This perhaps proves the immense importance of intensity of preference info
in the real world, and that all methods other than range, are dogmeat.
wds