Crystal Balling for Fun & Prophet
What if someone could use scientific research to divine the
sure winner of the 2004 election?
I love it when someone finds valid patterns where everyone
else sees chaos. I even love it when someone pretends to find valid
patterns where everyone else sees chaos. Today's
report
is courtesy
of Doc,
who sent
me a link
to research reported by Eric
Schulman, Ph.D, an astrophysicist who probably
understands math better than Doc and I combined. But Eric is also a humorist
who has authored a book, A
Briefer History of Time and a number
of articles spoofing the world of research. My favorite title: "The
History of the Universe in 200 Words or Less in 30 Languages or More in
Teeny Tiny Type"
Schulman has devised the Electability Formula, which looks
at candidates' years of experience in various public roles, and the actual
results of campaigns:
Electability = 4P - V - S + R + 9G + 95DCI + 95GEN + 95NUC,
where P is
the number of years the candidate served as President, V is the number
of years the candidate served as Vice President, S is the
number of years the candidate served as U. S. Senator, R is the
number of years the candidate served as U. S. Representative, and the
Boolean
variables DCI/GEN/NUC are 1 if the candidate served as Director
of Central Intelligence (e.g., George H. W. Bush), was a general officer
in the
United States Armed Forces (e.g., Dwight D. Eisenhower), or ordered
the combat use of nuclear weapons (e.g., Harry S. Truman), respectively.
Note that this is not necessarily a unique solution (i.e., we
stopped searching once we found a set of parameters that worked).
In each U.
S. Presidential election between 1932 and 2000, the candidate with
the higher electability won. |
Historically, Schulman contends, a President's
years in office has an advantage against most contenders (4 points per Pres.
year), unless he's running against a former Governor (9 points per Guv year)
or someone who has earned 95 bonus points. The 95 bonus points accrue to
anyone who ordered the use of a nuclear weapon
(N), has been a
General
(G), or was Director of the CIA (D). Such statistical gymnastics was
the only way Schulman could explain Truman's defeat of Dewey in 1948,
Eisenhower's wins against Adlai Stevenson or Dubya's dad against anyone.
You'll notice that
it's a negative to have been a Vice President or a Senator, which probably
comports with your personal view.
By jiggering the formula until it worked, Schulman has been
able to rationalize
the outcome of every presidential race since 1932. This is the kind of
thing that Wall Street's technical analysts do, so brokers can promise amateurs
they can beat pros in the stock market. Schulman's point is that his is a
bogus
indicator,
only true retrospectively. The question
is whether it will still be retrospectively valid in 2005.
Theoretical Musings, 2001
Americans seem to love experienced Governors who run for office,
and not Senators or VPs. In 2001, Schulman
noted, tongue firmly in cheek, that it would take a 4-term Governor to defeat
George W. Bush:
"An empirical formula is of little use if it cannot
predict future events. Should George W. Bush remain in office and run
for re-election in 2004, he would have an electability of 70. The Democrats,
however, could defeat him if they nominated James B. Hunt, Jr., the four-term
governor of North Carolina. But Republican strategists will no doubt
have read this article, too, and could respond by nominating four-term
South Dakota Governor William J. Janklow. Both candidates would have
electabilities of 144 (assuming they have no other relevant government
experience between now and 2004). Two candidates with tied electabilities
would surely lead to the closest U. S. Presidential election in 75 years."
|
Candidate
|
Year
|
Pres
|
VP
|
Sen
|
Rep
|
Gov (x9)
|
D/G/N?
(95)
|
Electabilty
|
|
George W. Bush
|
2004
|
4
|
0
|
0
|
0
|
6
|
|
70
|
|
James B. Hunt, Jr.
|
2004
|
0
|
0
|
0
|
0
|
16
|
|
144
|
|
William J. Janklow
|
2004
|
0
|
0
|
0
|
0
|
16
|
|
144
|
"This catastrophe could be avoided if President George
W. Bush orders the combat use of nuclear weapons before November of 2004,
in which case his electability would jump to 165, comfortably larger than
any of the possible Democratic candidates."
Outcome if George Bush Manages to
Nuke Somebody |
|
Candidate
|
Year
|
Pres
|
VP
|
Sen
|
Rep
|
Gov (x9)
|
D/G/N?
(95)
|
Electabilty
|
|
George W. Bush
|
2004
|
4
|
0
|
0
|
0
|
6
|
N
|
165
|
|
James B. Hunt, Jr.
|
2004
|
0
|
0
|
0
|
0
|
16
|
|
144
|
|
William J. Janklow
|
2004
|
0
|
0
|
0
|
0
|
16
|
|
144
|
2003 - Real News for a Real Election
We are indeed fortunate that Dr. Schulman updated his
research at the end of June. He applied his groundbreaking algorithm to the
field of Democratic hopefuls and concluded that, despite General Wesley Clark's
95 point advantage from his Generalship, Howard Dean is the best man to beat
Bush,
if the Dems
are smart enough to nominate him:
"George W. Bush will have an electability of 70 if he remains in office
through November of 2004. The Democrats could defeat him if they nominated
either former Governor Howard Dean (electability of 108) or retired General
Wesley Clark (electability of 95). Since the Democrats wish to regain the
White House, they will presumably nominate one of these men."
|
Candidate
|
Pres
|
VP
|
Sen
|
Rep
|
Gov (x9)
|
D/G/N?
(95)
|
Electabilty
|
|
Howard Dean
|
0
|
0
|
0
|
0
|
12
|
|
108
|
|
Wesley Clark
|
0
|
0
|
0
|
0
|
0
|
G |
95
|
|
George W. Bush
|
4
|
0
|
0
|
0
|
6
|
|
70
|
|
Bob Graham
|
0
|
0
|
-18
|
0
|
8
|
|
54
|
|
Richard Gephardt
|
0
|
0
|
0
|
28
|
0
|
|
28
|
|
Dennis Kucinich
|
0
|
0
|
0
|
8
|
0
|
|
8
|
|
Al Sharpton
|
0
|
0
|
0
|
0
|
0
|
|
0
|
|
John Edwards
|
0
|
0
|
-4
|
0
|
0
|
|
-4
|
|
Carol Moseley Braun
|
0
|
0
|
-6
|
0
|
0
|
|
-6
|
|
Joseph Lieberman
|
0
|
0
|
-16
|
0
|
0
|
|
-16
|
|
Bill Bradley
|
0
|
0
|
-18
|
0
|
0
|
|
-18
|
|
John Kerry
|
0
|
0
|
-20
|
0
|
0
|
|
-20
|
|
Joseph Biden
|
0
|
0
|
-32
|
0
|
0
|
|
-32
|
It's the Algorithm, Stupid
Don't pay attention to Dean's advantage in real-world politics – the
75,118 people signed up to go to a Meetup tonight, or the 258,452 people who
have registered at DeanforAmerica.com,
or the campaign's ability to raise a half a mill by posting a graphic on
their site or by Dean's appearance on the covers of Time,
Newsweek and
U.S. News. None
of that matters. Schulman's math tells us all we need to know: Howard Dean
will be
our next president and, with any luck,
General Wesley Clark will be his running mate.
Except for the Rove factor.
Be Afraid. Be Very Afraid.
Schulman notes one wild card:
"However, should President George W. Bush order the
combat use of nuclear weapons before November of 2004, his electability
would jump to 165, comfortably larger than any of the possible Democratic
candidates."
12:17:40 AM
|
|