Oy vey. Today I spent all day finishing up my master's report. It took a good chunk of time but it is now in its "defensible" format. Ok, I am so tired of computers now so here is the abstract for my paper. Enjoy:
***
Though the subject of player choice for fantasy football teams is extensively covered in magazines, the internet, and even specialized computer programs, there has been no serious discussion on what is the best fantasy football scoring system to use. This problem is pertinent because the value of an individual player for a fantasy football team depends greatly on the underlying scoring system used. To make things more complicated, the current standard scoring systems are not optimal; each one is either a bad predictor of the final game score or a bad predictor of the correct winner of a fantasy football game. The goal of this paper is to find an optimal scoring system that can accurately predict the correct game scores, predict the correct winner for a fantasy football game, and reward players in the fantasy game for events that happen in the real game that are not linked to direct scoring statistics. This will be accomplished by using a regression model to find proposed scoring systems and then by comparing them on two different validation tests to find which scoring system is optimal under basic constraints placed on the possible values for the scoring system.
***
You can get the full paper here, as well as all the data (if you're into that kind of thing ^_^ ).
Rock on.
No comments:
Post a Comment