College basketball stat websites should track player consistency. Start with tracking how well a player does during their non-conference, conference, and top-25 (50, 100, or just strength of opponent) match-ups and then combine them in some ratio and track it. This seems like a metric for predicting future player performance within reason. I suspect there's a trend there where a player has a good or bad game ratio based on the number of games they've played. 

Random related thoughts:

- Perhaps some players are consistently more reliable when it comes to playing bigger ranked teams?

- Does time of day impact player's performance?

- What's the likelihood that one player pulls the rest of the team down when they are having a bad game?

These all seem like possible factors in future performances or at least interesting points to study.