[EM] World Series and EC
Anthony Simmons
asimmons at krl.org
Sun Feb 3 20:46:48 PST 2002
>> From: Adam Tarr <atarr at purdue.edu>
>> Subject: Re: [EM] World Series and EC
>> But most importantly, I think the analogy is a poor one,
>> because we expect and desire this inconsistency in
>> baseball, and sports in general. Not only is the World
>> Series inconsistent due to the multi-game format (why not
>> just play one 60 inning game over a few days?) but scoring
>> is made erratic by the inning breakdown in games (why not
>> just give a team 27 outs in a row to each team?)
>> These erratic effects are by design. We say "may the best
>> team win" because, on some level, we are aware that they
>> may not. This uncertainty is part of the excitement of
>> sports: sports are filled with surprising heroes and
>> storybook last-second wins. This is the stuff of dreams
>> and it sells tickets.
After having thought about this some, it's occurred to me
this is an important observation. Margins of victory tend to
be much greater in the electoral college than in the popular
vote. Usually, anyway; it didn't seem to work out that way
in 2000. In elections, proportional representation is a
reasonable ideal. In sports, it's the opposite. We don't
want a compromise outcome in which all of the teams agree on
a proportional scheme that distributes the championship
evenly. One appealing situation is teams that are evenly
matched enough that we can't predict the winner, and a method
of scoring that makes a decisive victory possible anyway, by
amplifying small fluctuations in performance.
In elections, we want chance eliminated because it doesn't
reflect the underlying reality of popular choice. We want
other irregularities (elimination of popular candidates in
the first round of IRV, etc.) for the same reason. In
sports, we want small differences accentuated.
This reinforces the similar effect of districts (and their
counterpart, a series of games) in sports and elections.
Except that the effect is desirable in one case and a
hindrance in the other.
>> These are NOT properties I want in an election method. I
>> do not want any fringe, "Cinderella" candidate to be able
>> to capture the presidency on "any given November." I do
>> not want to see "hope spring eternal" for the American
>> Nazi party. I want boring, predictable elections where
>> the best candidate wins.
>> -Adam
More information about the Election-Methods
mailing list