This Is Basketball logo

Player Efficiency Rating (PER)

The standard, widely-recognised efficiency rating — shown across the platform as a reference, next to our own TIB Efficiency Rating, which stays our default.

PER, devised by John Hollinger, condenses a player's per-minute box-score production into a single number. Unlike a raw efficiency total, it is per-minute and pace-adjusted, then rescaled so that the average player in a given competition-season sits at exactly 15.0. That fixed baseline is what makes a PER comparable: 15 is league-average, ~18 is a solid starter, 20+ is an All-Star level, and 25+ is a standout.

How we compute it

We follow Hollinger's standard method: an unadjusted per-minute rating (uPER) built from a player's makes, assists, rebounds, steals, blocks, turnovers, fouls and free throws — each weighted by a league-wide “value of a possession” and rebound context.

That uPER is adjusted for team pace, then normalised so the minutes-weighted average across the whole competition-season equals 15. Every input comes from our own box scores.

How to read it
  • League average~15
  • Quality starter~18
  • All-Star level20+
  • Standout / MVP calibre25+
  • Historic single-season30+

PER vs TIB Efficiency

They answer different questions. TIB Efficiency is a per-game production score with our own weightings, a win/loss bonus and a competition-strength adjustment — it rewards who produced the most, night in and night out. PER is per-minute and rescaled to a fixed league average, so it isolates how efficient a player was with their time on court.

A high-minutes starter tends to top TIB Efficiency; an efficient role player can rank higher on PER. We show both so you can look at whichever fits the question — TIB Efficiency remains the platform default.

A note on small samples

PER was designed for long professional seasons. Over a short competition — a youth or national-team tournament of only a handful of games — it is inherently noisy, and a big performance in very few minutes can produce an extreme value. Our leaderboards apply a minimum-games qualifier for that reason, but treat a PER built on a small sample with appropriate caution.