I'm interested in exploring the effect of NBA team lineup changes on player prop markets.

Currently I bet spreads and MLs when the status of NBA players change before markets can adjust. This is profitable, but only for players who are "superstars" and only when changes are unexpected by the market which does not happen often. Figuring out which of these wagers to place isn't too difficult -- I keep some basic spreadsheets of historical line moves for key players as well as using Wins Above Replacement (WAR).

I'd like to expand my existing codebase to identify +EV bets on player proposition markets. The basic question is: If a player is unavailable or placed on a minutes restriction, which teammate benefits the most in terms of time played and other statistical categories? Simple depth charts likely won't be accurate enough. I'm wondering if there is some existing tool or model (possibly from the DFS world) that could assist in answering this question. The other option would be to collect raw historical data and use basic statistics to find correlations between one player being off the floor and another being used in his place. This could be done using game-by-game stats or more granular play-by-play data.

Any thoughts, suggestions, feedback?