This is a simple resume-based algorithm I developed for ranking college football teams. It just uses wins and losses.
- Preaseason rankings can bias a poll's effectiveness. To overcome this, this algorithm uses randomly generated preaseason rankings.
- In most polls a recent win or loss tends to be counted more strongly than earlier wins. This algorithm randomly shuffles the records of all games in order to alleviate this problem.
- Only wins and losses are counted in the algorithm (see below). Strength of schedule is built in, because winning against a team ranked higher in an iteration moves your team up.
See docs/about.md
for more information on how to
compile the program and how the algorithm works.
After 11 weeks of play in the 2017-2018 season: