I have an algorithm which takes as input sporting event outcomes and creates a rating of teams based on this.
My difficulty is the algorithm does not scale with time.
To give a simple example of the problem, assume a sport team rating method that rates a team based on their number of losses subtracted from their number of wins. (wins-losses). While for any given time it is easy to compare team's relative strength, it would not be possible to compare a team from one week of the season to the next as the score grows with time. (An easy fix for this rating method is of course to simply divide by the number of games played).
My algorithm is somewhat more complicated but the problem is the same. I wish to scale it so that the range of the rating is comparable for any timeframe.
The algorithm uses matrices and matrix math extensively and the Freelancer should be well versed in these.