
Originally Posted by
curious
I build a mathematical model of each team that includes factors like offensive efficiency, team contribution to defense, strength of schedule, defensive efficiency, regional rivalry, motivation, etc. This model then predicts how strong each team is if they played today. I then compare the numbers for the teams that are playing each other to see how much they would score against each other. That predicts the score for each game. I then equate the predicted score to both a spread and a moneyline. For any game where I think my spread or my moneyline are more accurate than the linesmakers and I think I have an edge, that is a play.
My system is really good at finding big dogs which win and the only reason they are big dogs is because the other team is being evaluated as better than they are because of some bias, usually publicity or popularity, so the line set by the linesmakers includes their bias.