Ok, so some of you may have seen
my previous diary where I published my letter to my winger friend who's a non-citizen and volunteering for the BC '04 reelection down in Florida as a lawyer and a Bosox fan. Basically it was an impassioned plea to stop working for BC. Here's a recent response I got in reply to my pointing out Kerry will win based on the Redskins effect.
that's the strongest argument you've ever made. of course, since the
red sox broke all curses with their derring-do, it also is unconvincing.
My response below the fold....
As I've explained to you before, the law of conservation of improbable events
clearly is against you on this one. The Redskins hypothesis favors
Kerry and it would be incredibly unlikely that, based on past
observation, it would be wrong. For it to do so would require an
improbable event of enormous magnitude, say, the President actually
admitting he lied to the American people about Iraq. Then maybe
you're right. But we know that won't happen. ;)
Just FYI, b/c I know your statistically-impaired brain wouldn't be up
to the task. Let us take the null hypothesis that the Redskin effect
is due to just chance alone. The alternative hypothesis, of course,
is that the Redskins disposition on the field is somehow inexorably
linked to the incumbant party's fortunes on election day, through a
yet-to-be determined mechanism. If we assume that the independent
probablity of a Redkins victory on the Sunday before election day of
presidential years is 50% (which certainly isn't true depending on the
opponent, but for the long time-frame included in this analysis and
the relative parity of NFL teams over the years, it's probably very
close), then the probabilty that this effect is due to chance alone is
0.5^n, where n= the number of observations. This is the same
probability of getting n heads or tails in a row. In the Redskins case
n=17. That makes the probability .0000076, or a 1 in a 100,000 chance
(if you round up).
So, this is incredibly unlikely to be due to chance alone. Now let's
look at the so-called curse of the Bambino. The Red Sox have been in
the series 4 times (previous to this year) since their ill-fated trade
of Babe Ruth and their last World Series victory in 1918. If we
assume the same probability distribution holds, that winning a world
series is no more likely than a 50% chance, then the probability that
the curse is due to chance alone is only 0.5^4 or a 1 in 16 chance.
That's a 6.25% chance, and well within the 5% chance we scientists
usually set the bar at for being statistically significant.
So to recap, Curse of the Bambino, NOT statistically significant.
Redskins win-loss record on Sunday before election day: incredibaly
statistically signficant.
Oh and as a refresher course:
WMD in Iraq: NO EVIDENCE
Al Qaeda and Saddaam in cahoots for 9/11: NO EVIDENCE
370 tons of explosives looted under the US's watch and now probably in
the hands of the terrorists: CAUGHT ON TAPE
What are they putting in the water down there that makes you guys see
black as white and true as false?
FWIW, my made up law of conservation of improbable events is as follows: For an improbable event to occur to an entity, the entity in question must commit an improbable action of equal or lesser probability.
In the case of the Bosox, breaking the curse (an improbable, but not incredibly improbable event) was offset by coming back from a 3-0 deficit in the ALCS (that's pretty improbable).
I'm fairly certain that this law is derivable from the 2nd law of thermodynamics. :)