Rebounding rates: why rebound percentage is fairer than rebounds per game
Rebounds per game depend on minutes, pace and how many shots get missed. Learn how rebound percentage corrects for all three, how it splits offense from defense, and what it still leaves out.
In 1960, Wilt Chamberlain grabbed 55 rebounds in a single game against the Boston Celtics, a record that still stands. Over his career he averaged about 23 a night, and Bill Russell, his great rival, averaged about 22. Those numbers are so far above anything seen in the modern game that they invite two reactions. One is to conclude that they were supernatural rebounders who outclass everyone since. The other is to suspect that something about the era inflated the totals. Both reactions contain some truth, and sorting out how much of each requires a better tool than rebounds per game.
The tool is rebound percentage: the share of the available rebounds a player grabbed while he was on the floor. It answers the question the raw count cannot, which is how well a player did with the opportunities he actually had. It is a small change in arithmetic with a large change in what you learn, and it is a good example of how rates can correct the biases of totals.
This article explains why rebounds per game are a poor yardstick, how rebound percentage is calculated, how it divides into offensive and defensive versions, what it shows about some of the great rebounders in history, and what even a good rate cannot tell you. It sits alongside the other rate-based measures on this site, including the ones described in the article on pace.
Why the per-game count misleads
A rebound is the recovery of a missed shot, so the number of rebounds available in a game equals the number of missed shots. That seems obvious, but the consequences are important. Anything that changes the number of missed shots changes the number of rebounds, whether or not the players get any better at rebounding. Three things matter most.
Pace. A game with many possessions has many shots and therefore many misses. A player on a team that plays fast has more rebounds available than the same player on a team that plays slowly. In the early 1960s, when teams played at a tempo far above anything seen later, the number of misses per game was enormous. The tempo of that era is discussed in the article on the Russell-Chamberlain era.
Shooting accuracy. Teams in that period also shot a lower percentage than modern teams, so a larger fraction of shots missed. More misses means more rebounds. The league's collective shooting has improved over the decades, partly because of better shot selection and training, and the improvement has taken rebounds out of the game. A rebounder today has fewer misses to compete for than a rebounder in 1962.
Minutes. A player on the floor for 40 minutes has more chances than one on the floor for 25. Per-game counts reward playing time, which is a legitimate contributor to a team's success but not the same thing as rebounding skill.
On top of these, there is a fourth factor that is easy to overlook: the way a team plays defense and rebounds as a group. If one teammate is a dominant rebounder, the others may get fewer rebounds than they otherwise would. Rebounds are a shared resource, and one player taking more reduces what is left for the rest.
For all these reasons, a raw rebounding total is a mixture of skill, opportunity and role. Rebound percentage tries to separate out the skill.
The calculation
The basic idea is to divide the rebounds a player collected by the rebounds that were available while he was on the floor. The formula for total rebound percentage is:
TRB% = 100 × (player rebounds × (team minutes ÷ 5)) ÷ (player minutes × (team rebounds + opponent rebounds))
Here is the reasoning in words. The team and its opponent together grab every rebound in the game, so the sum of the two teams' rebounds is the number of rebounds available. A player who is on the court for some fraction of the game faces only that fraction of the available rebounds. The factor of team minutes divided by five converts the team's total minutes into the length of a full game for one slot on the floor, so that the player's minutes can be compared with it. The result is a percentage of the rebounds available to him, estimated from the box score.
An example makes it concrete. Suppose a game has 90 total rebounds for the two teams combined, and a center plays 30 of the 48 minutes. He would have roughly 56 rebounds available during his time on the floor, which is 90 times 30 over 48. If he grabbed 12, his total rebound percentage would be about 21 percent. A teammate who grabbed 7 in 20 minutes would have about 37.5 rebounds available and a percentage of about 19 percent, a rate nearly as good in fewer minutes, which the raw totals of 12 and 7 would badly understate.
The figures are estimates, since the formula assumes that the available rebounds are spread evenly across the game and that each player's time on the floor is typical. Real games have stretches with more misses and stretches with fewer. But across a full season the estimate is reliable.
Offensive and defensive rebounding
Rebounding is not one skill but two, and the rates separate them.
Defensive rebound percentage is the share of the opponent's misses that a player secures. It reflects the ability to box out, to read the carom of a missed shot and to protect the ball after it comes down. For a defense, ending a possession with a rebound is the final part of a stop, and failing to do so gives the opponent another try.
Offensive rebound percentage is the share of the team's own misses a player recovers. It is a harder task: the offensive player is generally farther from the basket than the defenders and often must go through a boxed-out defender to get the ball. As a result, offensive rebound rates are much lower than defensive rates. In most eras, a team rebounds roughly a quarter to somewhat under a third of its own misses and the defense gets the rest, a split that reflects the positional advantage of the defense.
The distinction matters because the skills differ and because the strategic value differs. An offensive rebound creates a new possession, which is worth a great deal: a team with the ball under its own basket scores at a high rate. A defensive rebound ends the opponent's possession and starts your own. The trade-offs involved in sending players to the offensive glass versus getting back on defense are explored in the article on offensive rebounding versus getting back, and the defensive technique is covered in the article on defensive rebounding.
Splitting the rates also helps with the sorting of players by type. Many big men post high defensive rebound percentages because they are positioned near the basket on defense. Fewer post high offensive rebound percentages, because it requires energy, timing and a willingness to crash the boards at the expense of floor balance.
What the rates show about the greats
Rate statistics let us look at some famous rebounders with a fresh eye.
Dennis Rodman is the standout case. He led the league in rebounds per game for seven consecutive seasons in the 1990s, an extraordinary run, and he did it at a time when league-wide tempo and scoring were low. He was not a tall player by the standards of big men, and his edge came from preparation, positioning and relentless effort, as described in the article on Dennis Rodman. His rebound percentages are among the highest ever recorded for a long career. Because they are rates, they show that his dominance was not just a product of playing heavy minutes on a team with many misses. He took a bigger share of the available rebounds than almost anyone.
Chamberlain and Russell are the opposite case. Their totals were enormous, partly because the number of available rebounds in their era was enormous. Rate statistics reduce the apparent gap between them and the later greats, though they remain among the best ever. Sorting out their dominance by era is difficult because the data for rate statistics require full team and opponent totals, and the computations for the 1960s rely on estimates in some categories. Still, the rates show that they did not simply benefit from the tempo; they were outstanding relative to their time. Their careers are covered in the articles on Wilt Chamberlain and Bill Russell.
For more recent players, the same lesson applies in reverse. A modern center who averages 11 or 12 rebounds a night may be grabbing a larger share of the available misses than a player of the 1960s who averaged 20, because the number of misses available has fallen so much.
Why a rate is fairer
The reasons rebound percentage is more equitable than rebounds per game come down to four ideas.
- It controls for pace. Fast and slow teams are put on equal terms.
- It controls for shooting accuracy. A team that misses more than most gives its rebounders more to grab, and the rate accounts for that by using the actual number of misses.
- It controls for minutes. A starter who plays 36 minutes and a reserve who plays 16 are compared on the basis of the opportunities they had.
- It controls for the era. Because the denominator moves with the number of misses, the rate is roughly comparable across decades, subject to the data limitations.
The rate does not control for everything. It does not account for teammates, as discussed below. It does not account for the difficulty of the individual rebound. And it treats all minutes as equal, which is not true if a player is on the floor mostly in situations that favor or hinder rebounding.
What even a good rate leaves out
Several issues remain.
Teammate effects. Rebounds are shared. A player who grabs a high share of rebounds may be doing so because his teammates are leaving them to him. A defensive team that asks its guards to leak out early while the center collects everything will give the center a high rate and the guards low ones. That may be an efficient division of labor, but it makes the individual rates a function of the system. The reverse also happens: a center who boxes out his man so that a teammate can collect the rebound gets no credit in the figures. That type of work is a major part of the rebounding craft and is invisible in the box score.
Contested versus uncontested. Tracking data has made it possible to distinguish between a rebound grabbed in traffic and one that fell to a player alone. A high rate that comes mostly from uncontested rebounds is less impressive than one built on contested ones. Public data on this is limited and mostly confined to recent seasons.
Rebounds in the air. A player who tips the ball to a teammate or keeps it alive does something valuable that no statistic records.
The cost of the pursuit. A player who gambles for offensive rebounds may leave his team vulnerable to a fast break. A player who chases rebounds out of his area may leave his own man open. The rate says nothing about the costs.
Position. Players at different positions face different circumstances. Guards are typically farther from the basket and gather fewer rebounds. Comparing a guard's rate to a center's is not very informative; the right comparison is within a position, or against the team's needs.
Sample size. Rates stabilize reasonably quickly over a season because rebounds are frequent events, but short stretches can still mislead.
The point of the list is not that the rate is bad but that it is a description of one dimension of play. The article on analytics myths says something similar about all single-number summaries.
Team rebounding
Rates work for teams as well as for players. A team's offensive rebounding percentage is the share of its own misses that it recovers, and its defensive rebounding percentage is the share of its opponents' misses that it recovers. These are two of the numbers in the four factors, the framework that breaks a team's success into shooting, turnovers, rebounding and free throws. In that framework, rebounding is expressed in exactly this rate form because it is the only fair way to compare teams that play at different speeds and shoot different percentages.
At the team level, the relationship between the two kinds of rebounding is an important strategic choice. A team can prioritize the offensive glass, sending several players to crash and accepting the risk of giving up transition points, or it can prioritize getting back, conceding some second-chance points in exchange for defensive balance. Different eras and different coaches have chosen differently, and the choice shows up in the rates.
On the leaders page, you will typically find rebounds per game listed. Treat that as the starting point and look at the rate when you want to know who is really good at the craft. For teams on the standings page, the rate is the natural way to see which defenses end possessions and which offenses extend them.
How to read a rebounding line
A short routine will help.
- Look at the per-game number to see the workload.
- Look at the rebound percentage to see the skill relative to opportunity. As a rough guide, a percentage in the high teens is solid for a big man in a typical modern season, with the best well above that, but the numbers vary with era and position.
- Split offensive from defensive, since the two tell different stories.
- Consider the team context. Is the player a sole rebounder, or does he share the work?
- Compare within position and era, not across.
- Check minutes before drawing conclusions from a hot or cold stretch.
What to take away
Rebounds per game confuse a player's skill with the number of misses, the speed of the game and the minutes he plays. Rebound percentage divides by the opportunities, which makes it a much fairer measure and allows sensible comparisons across eras. Splitting it into offensive and defensive versions separates two different skills with different strategic meaning.
It still cannot capture boxing out for a teammate, the difficulty of the rebound or the system a team uses. Read it as the best available single measure of rebounding craft, then fill in the rest by watching how the work is shared. Chamberlain's 55 and Rodman's seven titles both look different, and more interesting, once you ask not how many rebounds a player collected but how many were there to collect.