Betting systems and strategies

For the past couple of months, I've been refining a system of mathematical models that estimates win probabilities for outcomes of sporting events in various leagues. I then use these results to determine value on moneylines and spreads. The betting strategy is ultimately built upon the predictive capabilities of crowd sourcing, aggregating data, and finding value.

I've developed strategies for NBA, NFL, MLB, NHL, NCAAF, NCAAB, WNBA, and CFL, and consider the following models/rankings in my system:

• TeamRankings (NBA, NFL, MLB, NCAAF, NCAAB)

• NumberFire (NBA, NFL, MLB, NCAAF, NCAAB)

• Jeff Sagarin's USA Today ratings (NBA, NFL, NHL, NCAAF, NCAAB)

• Log5 (based on teams' winning percentages) (NBA, MLB, NHL, WNBA)

• Pythagorean Expectation (originated by Bill James) (NBA, NFL, MLB, NHL, WNBA, CFL)

• KenPom (NCAAB only)

• Seven Overtimes (NCAAB only)

• ESPN Consensus data (NBA, NFL, MLB, NCAAF, NCAAB)

• Covers Consensus data (NBA, NFL, MLB, NHL, NCAAF, NCAAB, WNBA, CFL)

I then use the following criteria to determine my picks:

• For individual games and props, I only take value >= 2.5%

• For series futures, I only take value >= 5%

• For season futures, I only take value >= 10%

Finally, I grade these picks on a 1-5 star scale based on multiple criteria, including public consensus, whether the models strongly agree with one another, etc. Starting August 3, I'll begin posting my daily picks on the blog, as well as tracking them on my Covers account.