[wyowolf]

I sure as hell hope he is wrong. with the polls all over the place and being so close :(

Someone is going to look really stupid Wednesday, I sure hope its not us.

[pieceofthepuzzle]

The problem with applying the same mathematical methodology that one uses to make predictions in baseball to predict elections is that they are based on completely different types of data sets.

In baseball you have hard statistics, based on a player's past performance. If the statistics state that a player reached base 30% of at bats, that's what it is. It's not 30% with a margin of error of 3%, or 30% when the model uses 41% Red Sox fans vs. 30% Rangers fans. It's 30%, period.

Polling, on the other hand, always has a margin of error, and depends heavily on the models used in the particular polls being evaluated.

In short, there is significantly more variability and vagary in the input data for political mathematical prognostication than in prognostication in baseball.

If it were that straightforward, and like calculating loads on a bridge, they could just come up with the equivalent of Turbo Tax that would automatically input the polling data from each organization and spit out the likely winner. No need for commentators or prognosticators - just the ‘Turbo Election’ program.

It's not that simple.