Let's check it out.

Bergman's Fargo rating this morning is 791; Busty's is 790. A statistical wash.

**Effectively 1:1 odds for each to win a game by the Fargo ratio calculation**. (ratio = 2^(1/100) = 1.007)

*However*, in the two true 8-8 games (8 ball and one pocket) they just played, they finished like this:

__8 ball__: Bergman 25 / Busty 21 (expected = 23 / 23)

__one pocket__: Busty 8 / Bergman 2 (expected = 5 / 5)

The even money odds for each based on

*the empirical results* would have Bergman giving 1.19 : 1 on the money in 8 ball and Busty giving 4 : 1

on the money in one pocket.

The real world results differed significantly from the expected results by Fargo

**(8.7% error and 60% error)**. Although the 8 ball predictions came somewhat close if 5% is an acceptable error, the one pocket difference is massive.

Moral of the story... it's more-or-less

**a ***crap shoot *to rely on statistics - no matter how technically correct the method - to predict the outcome of single events, or a series of them. The same problems exist when converting ball spots to money odds, especially when the calculation tends to compound errors like this: [ball spot] > < [Fargo ratio] > [money odds].