We start by constructing simple systems that show consistent long-run results and make sense form what we know about sports. We add refinements and validate our models with real-time out-of-sample testing. We monitor and adjust all our systems after their adoption.  

Our systems are a blend of fundamental and situational factors. They often exploit mean reversion as in team bounce-backs or let downs. They also exploit public and oddsmaker behavioral biases. These include myopia, recency, and familiarity. 

.