Many people cite numbers per game. There is some merit in that, after all, wins and losses are tallied at the end of games. Thus, it makes sense to examine someone’s average production over the course of a game.
But what happens when one of those games has 120 possessions and another 80 possessions? Consider the following example:
Points Per Game | Possessions Played | |
---|---|---|
Player A | 25 | 120 |
Player B | 20 | 80 |
Player A scores more points in a game, but he took 50% more possessions to do so. Since there is nothing inherently better about playing slow or fast – each team will have an equal number of possessions – it makes little sense to ignore number of possessions in a game. Just like it would make no sense to compare statistics from a 63 minute game to a 48 minute game to a 24 minute half as if they were representative of the same level of production.
That 120 possession game is going to have more scoring, more rebounds, more of everything. Just like a box score at the end of a game versus a box score at halftime. If Michael Jordan only plays 50 possessions in a game he won’t rank in the top-20 in scoring, but he might be scoring more frequently than anyone in the league. I’m interested in what a player does per possession, or over some normalized amount of possessions.
[…] Basketball-Reference doesn’t have a pace-adjusted scoring metric. I normalize most of my stats to an estimated 75 possessions played, which for points produces a “scoring rate,” per […]