Re: Bush/Kerry optimal wagers

From: John Conover <>
Subject: Re: Bush/Kerry optimal wagers
Date: 1 Nov 2004 01:04:32 -0000

There are two days to the election, and although the numbers on the
popular vote between Bush and Kerry have closed, the probability of
Bush winning has not changed-in fact, it is slightly better, (the
reason is that there is less time for the fractal to change.)

There is about a 2% difference between Bush and Kerry in the popular
vote, or about 1% if its regarded as a zero-sum game, (in the lay
sense, ignoring the formalization of utility in zero-sum games; what
one wins, the other loses.) With 2 days left, the fractal could be
moved by about sqrt (2) = 1.414%, and the standard deviation
corresponding to 1 / 1.414 = 0.707 is about a 76% chance of Bush

However, (and its a big however,) the 18-29 year old vote could sway
the election in Kerry's favor-if there is adequate turn out; they are
not counted in any of the polls. Fortunately, this age demographic is
Internet savvy, and an assessment can be made from their 'blogs.

The rel(1) program at was originally
developed to monitor "BizWire," and other press release agencies for
equity related information, and feed it-nearly real time-to
tsinvest(1) to give it a "heads up."

After seeding wgetrel(1) with a quick search of Google, (wgetrel(1)
will "read" the HTML search presentation from Google,) and letting it
run for several days-searching the Internet for political 'blogs-about
65% of the 'blogger posters favor Kerry. (All I did was to pick out a
thousand favorable Bush messages and messages from a thousand Kerry
supporters, strip the phrases from the 'blogs postings, and feed it
back to rel(1) as a series of n-grams, in an iterative search

There will probably be about a hundred million voters turn out on
November 2, and if 2,000,000 / 0.65 = 3 million of them are in the
18-29 year old demographic, it would nullify Bush's lead.


BTW, if the younger demographic voters determine the election this
year, it is very feasible that there could be a third party with
enough resources to be a threat to the Democrats and Republicans in
2008. A lot of the n-grams indicated no favor of the Republicans or
Democrats, only a lesser-of-two-evils on some single specific issue.
(A viable third party may not be in the best long-term interest do to
Kenneth Arrow's so-called "Impossibility Theorem," BTW, the ranking of
priorities is intransitive between 3, or more, ideologies, and there
can never be a mechanism by which cooperation/compromise can be

It is probably a virtual certainty that there will be voting legal
challenges. Using Bush to win the electoral vote with 55%, and the
popular vote by 60%, the chances for Bush to win the electoral vote,
and not the popular, is 0.55 * 0.4 = 0.22. Likewise for Kerry, it is
0.45 * 0.4 = 0.18. Taken together, 0.22 + 0.18 = 40%, which is the
chance that whoever wins either the popular or electoral vote will not
win the other.


John Conover,,

Copyright © 2004 John Conover, All Rights Reserved.
Last modified: Sun Oct 31 17:16:47 PST 2004 $Id: 041031171327.25041.html,v 1.0 2004/11/01 01:19:56 conover Exp $
Valid HTML 4.0!