I misworded that question and you're missing the point. What a player produces per minute is what they did produce, divided by the number of minutes that they play. They are actual bottom-line numbers, and they're more reflective of performance than per game stats. Find me an advanced basketball metric that isn't in the box score that doesn't involve minutes played or pace and we can talk. If your friend is chucking up several shots (3? 4?) and making one, a 25% shooting percentage clearly doesn't warrant more playing time than he was getting.
I think that you're alluding to some biases that per minute stats can have, which are that a player who plays 2 minutes a game is generally playing against weaker opposition, and that players who play more minutes are generally better, and should get recognized for that. While it's true that you shouldn't compare the per minute statistics of a player who is playing 2 minutes to a player who is playing 35, that kind of reality check should come along with any statistical analysis.
A good example is Moriah Jefferson. She's playing 28 minutes per game, not because she's not crucial to the offense, but because UConn has blown out so many of the teams that they've played. When comparing her stats to Jewell Loyd (who's up at 32 minutes because her team has needed her to play for longer), do you think her production should get docked because she's playing fewer minutes? If that's the case, what's the value of per game stats at all, why not just look at totals on the season?
