We all know the larger the spread the more likely a game is to be blown out. This means starters will play less minutes while bench players will play more.
What would be a good way to factor this into a simple NBA points prop model? Does anyone have historical data that may support adjusting X percent of minutes based on every point in the spread?