Darin Sunley wrote:
>
> >Dave Sill wrote:
>
> >Harry Browne needs a vote for him less than Gore needs a vote against him.
>
> The way this is phrased suddenly caused an election analogue of Baye's
> Theorem to flash into my head.
>
snip
>
> I wonder what the analagous effect of this assymmetry on political calculus
> is?
This can be measured by measuring the size and strength of positive and negative
opinions of each candidate in a poll. The real numbers that political pollsters
look at is not the pap that is published in the media. What is of importance are
things like how 'firm' the support for a candidate versus his opponent, as well
as the average vector of undecideds making up their minds. Firmness is measured
by asking multiple choice questions when asking who a person will vote for,
asking if they a)possibly, b) probably, or c) definitely will vote for a
candidate. This firmness can also be measured in the negative opinion of a
candidate. A candidate may have 20% firm support, 25% soft support, but 35% firm
opposition, 10% soft opposition, and 10% undecided (that 45% firm opposition
does not neccessarily translate to 45% firm support for the opposing candidate).
A professional pollster then looks at trends in these various variables, coming
up with derivative trajectories for each, which he integrates together to make
one full caculation of who will win what percent of the vote on any given day.
What is more difficult, and is considered more of an art among pollsters, is
measuring how outside events can change this dynamic: i.e. how candidates do in
a debate, if new scandals about candidates surface, how the media spins each
candidate to the public, as well as world events and economic trends. Dealing
with this is very much like Hari Seldon's Science of Psychohistory.
This archive was generated by hypermail 2b30 : Mon May 28 2001 - 09:50:16 MDT