Monday Night Football showed some very weird stats last night. Since 1940, the Redskins have either lost or won their monday night games and the encumbent for president has either won or lost right along with them!

In other words, if they would have won last night, Gore should win next week. Since they lost, if they stay with this pattern (didn't one of the conspiracy freaks say something about a simple pattern for see who would be president? hmmmmm....) It means Bush will win.

The pattern was something like 16-0. Will they make 17?

-- eiry (sort@strange.huh?), October 31, 2000

Answers

oh and for the record, I don't believe in conspiracies, I just found this statistic irresistable.

-- i cant spell or type today (sort@strange.huh?), October 31, 2000.

I'm fairly certain there was no MNF during the 1940's. :-)

Just a hunch, but I'm pretty sure about this one......

Deano

-- Deano (deano@luvthebeach.com), October 31, 2000.

No Monday night games before 1970. Dandy Don retired after the 1968 season and took a year off. He considered rejoining the Cowboys in 1970, but received the MNF offer and the rest was history.

-- Bingo1 (howe9@shentel.net), October 31, 2000.

Another sterling lesson in the difference between correlation and causation.

-- Brian McLaughlin (brianm@ims.com), October 31, 2000.

I understand causation.

Seems prior to one MNF broadcast Howard Cosell drank a great deal of vodka or similar spirit, tossed his cookies all over Dandy Don's cowboy boots. Howard was taken off the air at halftime.

Causation of his cookie toss was the vodka. As to correlation I have no clue. Anybody know where I can buy one cheaply?

-- Bingo1 (howe9@shentel.net), October 31, 2000.

Bingo, sorry to leave you stranded without a clue.

In statistics, you may correlate two sets of data a number of different ways. One set of data may go up when the other goes up, or one goes up when the other goes down, or one goes up and down exactly three days after the other goes up or down. And so on. These are correlations. You can quantify these correlations to see if they are "statistically signifigant".

However, you can do odd and wonderful things with correlations.

For example, if you find that, in years when the field mouse population in Zambia goes up, the casinos in Las Vegas record an usually high number of slot machine jackpots in November, then you have found a statistical correlation. It might even be a very strong, or complete correlation stretching over many decades.

But you can be damn well certain that the mice in Zambia and the slots in Vegas have no causal connection whatsoever.

-- Brian McLaughlin (brianm@ims.com), October 31, 2000.

Much appreciated, Brian. There are huge gaps in my education and my ability to remember what I was present and accounted (though not accountable) for. Somehow the causation/correlation thingee passed me by or slipped out a hole in the sieve.

Pretty basic stuff. How embarraskin!

Rich

-- Bingo1 (howe9@shentel.net), October 31, 2000.

And although the source of the following article has, shall we say, variable reliability, here is an even more intriguing example of a correlation with no obvious causative link:

http://www.yowusa.com/Archive/October2000/Tecumseh2/tecumseh2.html

Excellent explanation BTW, Brian.

-- (Lewis@NarniaGeneral.Delivery), November 01, 2000.