Quote:
Originally Posted by Vegas Vic
I wanted to update this for anyone who is following. I've given some thought to this, and I'm going to change the process slightly. Up to this point, I've just been using a one game simulation to determine the winner of each game. While this is obviously more expedient, it yields a significant amount of variance.
What I'm going to do is run 101 simulations of each matchup, with the winner being determined on who wins the majority of those simulations. Then, the actual game of record will be the last game of those simulations that has that particular team winning.
|
Fun logic here.
We hate the "scourge of the BCS" so we want to find out the true champion on the playing field, the way it's supposed to be.
...a little time passes...
OMG this is way too random, the good teams are too bound to lose in a one-off matchup, we need to fix that "variance."
It's almost like forcing the #1 team in the country (that being earned over a full season) a mere one-and-done matchup with the #8 or #9 team is an insufficient way to reward their full year's worth of credentials. Hmm. A tiny bit less scourge-y perhaps.