So this has been bugging me ever since ESPN started their ridiculous FPI rankings and stats, but there have been a few people now (ESPN being most prominent and dedicated to it) who roll out their win probability statistics as if it's a real measure with some sort of legitimate math/evidence behind it. ESPN has now doubled down with their site fivethirtyeight.com by putting together probability measures of getting into the playoffs. Those measures are in the link here:

http://fivethirtyeight.com/features/the-week-in-college-football-a-high-...

Now I'm no math major, far from it actually, but I'm fairly certain that there is absolutely no way for anyone to calculate win probability in a manner that would be even the least bit legit. Furthermore, I'm more certain there is no way they can predict probability of playoff inclusion for a playoff determined entirely off of a 12 member committee that has proven to be fairly fluid in the past.

Seeing as both win probability and playoff inclusion are nearly exclusively human-determined with literally thousands of potential outcomes, I find the inclusion of things like FPI as a general metric alongside actual, proven metrics (wins and losses, winning percentage, etc.) to be incredibly irresponsible (I'm similarly skeptical of all the advanced statistics out there in other sports but those at least seem more realistic).

So my point in posting this is as follows: could one or more of the math genius Dubbers please weigh in and try to explain in dumb persons terms whether these measures are legit? This seems incredibly like numbers are just being pulled out of thin air.

Note: this is not a bash ESPN thread. There are plenty others around here to do that. I'm just trying to get educated.