|
 Prediction
Markets Help Find
Lost Subs... And
They Forecast
Election Outcomes
Too
ABC's
20/20 this past
Friday reported on
how prediction
markets are used to
determine the
probably outcome of
a political election
with a focus on
Ireland's online
gambling house,
Intrade.
The
news mag used an
interesting
comparison in its
report, citing how
large groups of
experts were
correctly able to
locate a lost sub by
simply predicting
where it might be
located. The
majority opinion
proved the most
accurate.
The 20/20 report
referred to this as
"Wisdom of Crowds".
James Surowiecki
explained why
prediction markets
are so accurate in
his book "The Wisdom
of Crowds."
We have heard bad
things about crowds.
They're
uncontrollable and
prone to mob
psychology. But
under certain
conditions,
Surowiecki argued,
crowds can be wise.
"Even though no one
person in the crowd
knows everything,"
Surowiecki said, "if
the crowd is big
enough and it's
diverse enough, you
just have access to
so much more
knowledge than you
do if you ask one
expert or even a
really small team of
experts."
You can see this on
"Who Wants to Be a
Millionaire?" When a
contestant is
stumped, they can
call an expert,
usually a smart
friend, or they can
poll the audience.
"The experts, they
do pretty well. They
get the answers
right about
two-thirds of the
time," Surowiecki
said. "But the
audience gets the
answer right 91
percent of the
time."
The wisdom of crowds
isn't limited to
game show audiences.
The Navy learned
about this principle
40 years ago when
the nuclear attack
submarine Scorpion
became lost
somewhere in the
North Atlantic. The
search went on
fruitlessly until a
Navy man named John
Craven assembled a
large group of
people, from
deckhands to
captains, and asked
them many different
questions that would
help predict the
location of the lost
submarine. He didn't
just ask them what
they thought, he
asked to them to bet
on it.
"He took all those
bets, and put 'em
all together, and
what he finished
with was this kind
of map of the ocean
floor," Surowiecki
said. "And on this
map there was this
one spot where his
team said, 'This is
where we think the
submarine is most
likely to be.' It
wasn't anywhere near
where the Navy had
been looking … and
they found it."
The submarine was
just 220 yards away
from where the bets
said it would be.
More recently,
officials in the
Department of
Defense wanted to
use the same
principle to set up
a market that would
predict where and
when the next
terrorist attack
would take place,
but when the news
broke there was an
enormous backlash.
"The idea of a
federal betting
parlor on atrocities
of terrorism is
ridiculous and it's
grotesque," Sen. Ron
Wyden, D-Ore., said
at the time.
The Department of
Defense quickly
backed away from the
idea and Surowiecki
thought that was a
mistake.
"These are
potentially
tremendously useful
tools for improving
our national
security," he said.
"It's much more
egregious not to use
them than to use
them."
----
Payton O'Brien,
Gambling911.com
Senior Editor
POBrien@CostiganMedia.com
Originally published
May 12, 2008 10:51
am EST
|
