So this week I finally got the ranking program converted to something I can understand a little more. I've been tweaking it here and there, using last year's starting point/games as a test, and I've gotta say, there isn't much that can be changed to make things end up much better.
So far, the best I can do is get the program to pick 9 more correct games on the entire season, but that ends up with a very inaccurate (imo) final ranking. Here's what that top 15 would look like:
Additionally, no matter what values I change in the algorithm, it still never correctly guesses either of the UIL state championship games. In fact, none of the picks from week 14 onwards change. The spreads change, sure, but not the predicted winners.
I'm going to keep playing with this thing but the point is, I think @granger has thing thing nailed down pretty darn good!
So far, the best I can do is get the program to pick 9 more correct games on the entire season, but that ends up with a very inaccurate (imo) final ranking. Here's what that top 15 would look like:
Additionally, no matter what values I change in the algorithm, it still never correctly guesses either of the UIL state championship games. In fact, none of the picks from week 14 onwards change. The spreads change, sure, but not the predicted winners.
I'm going to keep playing with this thing but the point is, I think @granger has thing thing nailed down pretty darn good!
