Adjusting the Exit Poll.

[LostAngeles]

It's funny the things you come across while reading the Tuesday Morning Quarterback on www.nfl.com

Take this:

...
Speaking of exit polls, John Martin pointed out this unintentionally hilarious Washington Post article which declares, "After the survey is completed and the votes are counted, the exit poll results are adjusted to reflect the actual vote, which in theory improves the accuracy of all the exit poll results, including the breakdown of the vote by age, gender and other characteristics." So after the actual vote is known, the exit poll is adjusted to make it appear accurate all along! Isn't this the same as waiting until the game is over, then "adjusting" your prediction of the final score in order to claim you predicted the final score? Martin also lauded the election of TMQ's estimable Colorado College classmate Ken Salazar as a United States senator from Colorado.
...

There's a link there, but it requires registration. I think TMQ stated my feelings on this one best with

...
So after the actual vote is known, the exit poll is adjusted to make it appear accurate all along! Isn't this the same as waiting until the game is over, then "adjusting" your prediction of the final score in order to claim you predicted the final score?
...

Remind us of anyone we know? (coughLucianarchycough1inChristcough)

ETA: Here's the link to the column. It's down at the bottom, but take a moment to read the exchange between Terrell Owens and Ray Lewis.

http://www.nfl.com/news/story/7871001

 

[aerocontrols]

Yes, it's the equivalent of doing that, except that exit polls aren't primarily conducted to determine the winner.

Methinks the fine political analysts at nfl.com don't understand the purpose of exit polls.