The goal of this is to eventually predict outcomes of games. However, for now it also will rate teams in a league based on certain features.

For the future I would like to add the following features

• more sports leagues.
  • currently this only works with the NFL. However if I can find some API's similar to nflGame then I can get all of the data I need
• neural netowrk capability so that it will be able to predict
• ability to save the pandas representation of my ratings to csv
• lint the code and add comments and argument descriptions

If you run this right now, it will create an object called r which is the Rater class. I will probably be changing some of this around to be more accureate to the classes purpose once I get more work done.

#RATER

a simple rater can look like this

from LeagueRatings import Rater
rater = Rater('nfl')

with just that code you will then have the ability to get the overall rankings with default features and weights, if you would like to modify the weights before the code runs, you can do the following.

rater = Rater(start=False)
rater.weights = new_weights #some list of weights
rater.get_ratings()

Right now the main functionality here is to rate the various teams in a league. Currently I am using 9 features

• op = overall performance. Basically total wins
• hp = home performance. Number of wins at home
• ap = away performance. Number of win on the road
• pf = points for. Total points scored
• pa = points allowed. Total points allowed
• hpf = home points for. Total points scored at home
• hpa = home points allowed. Total points allowed at home
• apf = away points for. Total points scored on the road
• apa = away points allowed. Total number of points allowed on the road

I assign weights to each feature, and then apply those weights to each team's stat and sum them.

for the data below the normalized weights are as shown. They are in the same order as the features are listed above.

rater.weights = array([ 0.16666667,  0.05555556,  0.16666667,  0.11111111,  0.11111111,
        0.08333333,  0.11111111,  0.11111111,  0.08333333])

As of week 7 through the NFL season this is how the ranking shows.

rater.get_team_ranks(do_print=True)

You can also see the output of where each team ranks on each feature

r.print_table()