Player efficiency rating (PER) is a single number that tries to sum up everything a basketball player does per minute on the court, points, rebounds, assists, steals, blocks, shots made and missed, turnovers, all rolled into one stat pegged to a league average of 15.0. John Hollinger built it, and it’s still one of the most searched advanced stats in basketball because it promises a shortcut: one number to rank players.
That promise is only half true. PER is useful for a quick gut check on a player’s overall production. It’s a bad tool for picking a specific prop bet, and this article explains why, along with what actually works better for that job.
What is player efficiency rating?
Player efficiency rating is a per-minute rating of a player’s statistical production, adjusted so the league average sits at 15.0. Anything above 15 is considered above average, anything below is below average. Hollinger developed it while writing for ESPN, and it’s now a standard stat on sites like Basketball-Reference.
The formula is long and not worth memorizing. It adds up positive contributions (points, rebounds, assists, steals, blocks), subtracts negative ones (missed shots, turnovers, fouls), then normalizes for pace and minutes played. You don’t need to do the math yourself. You just need to know what the number is telling you and what it’s leaving out.
- Above 20: All-Star level production
- 15: league average, roughly a solid rotation player
- Below 13: below-average per-minute output
- 25+: historically great, MVP-caliber seasons
How is PER actually calculated?
PER starts with a “unadjusted PER” that credits a player for makes, assists, steals, blocks and rebounds, then debits misses, turnovers and fouls. That raw number gets adjusted for team pace (fast teams generate more counting stats) and normalized so the league average always lands at 15.0, no matter the season.
The key detail bettors miss: PER is a rate stat, not a volume stat. A player who goes 4-for-5 shooting in 18 minutes off the bench can post a monster PER for that game while contributing far less total production than a starter who plays 34 minutes and shoots 9-for-20. That distinction matters a lot when you’re trying to project points or rebounds for a specific game, not rank a season.
What does PER get wrong for prop betting?
PER tells you a player was efficient. It does not tell you what he’ll do tonight, against this defense, in this role. That’s the gap that matters if you’re betting points, rebounds or assists lines.
Three specific blind spots:
- No defensive matchup context. PER doesn’t know if tonight’s opponent is bottom-5 or top-5 at defending a player’s position.
- No usage or role signal. A high PER from garbage-time minutes reads the same as a high PER from clutch minutes in the formula.
- No recent form. PER is typically reported as a season average. It smooths over the last five games, which is exactly the window a prop bettor cares about most.
Key takeaway: PER answers “how good is this player overall,” not “will he clear this line tonight.” Those are different questions, and prop betting is built entirely around the second one.
What should bettors use instead?
For prop research, matchup-specific and recent-form data beats a season-long efficiency number every time. That means looking at how a defense ranks against a given position, how a player’s last 5 to 15 games actually looked, and how consistent his production has been game to game, not just his per-minute average across 70+ games.
StatsBench’s Prop Finder builds around that idea instead of a single efficiency score. It surfaces hit rate, a consistency grade, and defensive rankings by position side by side, so you can see whether a player’s line makes sense against tonight’s specific opponent, not just against the league average.
As an example of what that data actually looks like in practice: a StatsBench cheatsheet snapshot from a July 2026 Phillies vs. Dodgers slate showed Bryson Stott hitting his hits over 0.5 line in 6 of his last 10 games (60%), with total bases over 0.5 also landing at 60% (6/10) and a modest 0.9% edge at the best available price. Kyle Schwarber’s total bases over 0.5 sat at the same 60% hit rate with a 0.9% edge. None of that shows up in a PER number, because PER isn’t built to answer prop-level questions. It’s built to rank seasons.
| Metric | What it tells you | Good for prop betting? |
|---|---|---|
| PER | Overall per-minute efficiency, season-long | No, too broad |
| Hit rate (L10) | How often a player cleared a specific line recently | Yes |
| Consistency grade | How stable that production has been | Yes |
| Defensive rank vs. position | How tonight’s opponent defends that stat | Yes |
Does PER matter at all for props?
Yes, but as background context, not a betting signal. A player with a PER over 20 is a legitimate offensive engine, and that matters when you’re deciding whether a points or PRA line is even in the right neighborhood. It just can’t tell you whether tonight’s specific number is priced right.
Think of PER as a sanity check. If a role player suddenly has a monster points line, checking his season PER can tell you if the market moved for a real reason (an injury opening up minutes) or if the number just looks off. From there, the actual betting decision comes down to matchup, recent form and price, which is where tools built for props (not season stats) do the heavy lifting.
How do sportsbooks use efficiency stats when setting lines?
Books use their own internal models, and PER isn’t the driver. Their lines already bake in pace, matchup, injury news and role changes, updated constantly across the books. That’s exactly why a stat like PER, which lags and averages everything out, is a weak standalone tool for finding value against a line that’s already accounting for tonight’s context.
The real opportunity for bettors isn’t outsmarting the market with a smarter season-long stat. It’s finding the specific books that haven’t adjusted yet, or the sportsbook offering a soft number relative to the sharp consensus. That’s the job StatsBench’s Positive EV Scanner does, recomputing fair prices from sharp books roughly every minute and flagging where a soft book hasn’t caught up. It’s a different question than “who’s efficient,” and it’s the one that actually moves the needle on a bet.
Where PER fits alongside other basketball stats
PER is one entry in a longer list of advanced basketball metrics, alongside things like true shooting percentage, usage rate, and box plus/minus. If you’re building out prop knowledge more broadly, our guide on NBA usage rate covers a related metric that’s actually more useful for prop research, since it flags who’s about to see a bigger role before the market fully adjusts. And if you want the full picture on how to research a slate rather than lean on one number, the NBA player props strategy guide walks through the whole process.
For a refresher on foundational betting numbers in general, our Betting Statistics 101 post is a good starting point if terms like hit rate or consistency grade are new to you.
Bottom line
PER is a fine way to answer “who’s having a good season.” It’s the wrong tool to answer “should I bet this line tonight.” Those are different questions, and mixing them up is a common mistake for newer bettors leaning on advanced stats they picked up from a stat sheet rather than a betting context.
Betting on player props means betting on variance over a real sample, not a guarantee. Keep bets within your means, treat it as entertainment, and remember the standard responsible-gambling resources (1-800-GAMBLER) exist if betting stops being fun. You have to be 21+ where sports betting is legal to play.
If you want a starting point that already blends recent form, matchups and consistency instead of a single season average, grab the free StatsBench cheatsheet. It’s free, updated regularly, and built for exactly the question PER can’t answer: what’s likely to happen tonight.
Frequently Asked Questions
What is a good PER in basketball?
League average is set at 15.0. Anything above 20 is considered All-Star level, and anything above 25 is historically elite, MVP-caliber production. Below 13 is generally viewed as below average.
Who invented player efficiency rating?
John Hollinger developed PER while working as a basketball writer, and it became a widely used advanced stat after he brought it to ESPN. It’s now a standard metric on sites like Basketball-Reference.
Is PER a good stat for betting player props?
Not really. PER is a season-long, per-minute average, so it doesn’t account for tonight’s matchup, recent form, or role changes, all of which matter far more for a specific prop line than an overall efficiency score.
What’s a better stat than PER for props?
Recent hit rate on a specific line, a consistency grade, and how the opponent defends that position tend to matter more. These give you context about tonight’s game instead of a season-wide average.
Does a high PER mean a player will score more tonight?
Not necessarily. PER reflects overall efficiency across a season, not tonight’s matchup or workload. A player can have a high PER and still face a tough defensive matchup that limits his production for one game.