Quote:
Originally Posted by RacersEdge
just happened to tell my boss about this - ugh - he said there is too much randomness in games to use probabilities....punt.
this has really been shocking to me - otherwise, in this case, he's a pretty smart guy. i didn't realize how rare it is to think about things from a probability perspective - and how much people would resist it even when explained to them.
I think it'd be interesting to calculate the probability by using variances of each statistic. Obviously, there are going to be more punting data available than 4th and 2.
In other words, I think your boss kind of points out something very important when trying to calculate this answer. The variance could be pretty large to give us a +/-X where X is too big to make this mathematical calculation anything but significant.
Furthermore, the stats would obviously change based upon being on the road or not, indoor or outdoor, patriots or colts etc etc.
I'm looking forward to future information regarding this matter and possibly someone calculating it in a better, more accurate, fashion.
I just don't know if the figures calculated can even be considered "ball park" figures because they fail to recognize the uncertainty.