Inside the Numbers - Offensive Efficiency
By Michael Zarren
Can a team become significantly better offensively without scoring any more points per game? This month's column shows how this year's Celtics team, post-trade, is doing exactly that.
You frequently hear on TV that a team is poor offensively "because they only score 95 points per game." But is a team that scores fewer points per game than some other teams do really a bad offensive team? For example, imagine a team (we'll call them "Slow-Pace Team") that scores every time it gets the ball, but uses up the entire 24 second shot clock to do so. Now imagine another team ("Fast-Break Team") that scores half the time it gets the ball, but uses only 8 seconds each time down the court. Fast-Break team might score more points in a game, but only because of the quick pace of the game. Of course, Fast-Break Team's opponents would also get more possessions, and would probably score more total points due to the game's fast pace.
If Fast-Break Team and Slow-Pace Team each played an opponent that scored two-thirds of the time it got the ball, Slow-Pace Team (who scores every time down the court) would beat the opponent, and Fast-Break Team (who only scores half the time it gets the ball) would lose -- even though Fast-Break Team would score more total points against the opponent than Slow-Pace Team would. So would you really call Fast-Break Team, which scores only half the time, a "better" offensive team than Slow-Pace Team, which plays more slowly but which scores every single time down the court?
Most basketball people would agree that a better offense is one that scores the most points each time down the floor, regardless of the game's pace. To account for differences in the speed with which different teams play, Celtics coaches judge an offense's effectiveness by the "points per possession" ("PPP") stat, rather than just looking at the number of points a team scores per game. PPP measures how many points a team scores, on average, between the time it gets the ball and the time the other team gets the ball back.
We use scorers' table data to calculate this stat, but those of you who are math geeks can use this formula that estimates the number of possessions any team has had. The rest of you can trust that we know each time the ball changes hands, and that we divide our total points by the number of possessions, and then multiply by 100. This equation produces the ranking of true offensive efficiency in the table at right.
As you can see, adjusting for the speed of each team's game can reveal some good offenses that aren't very speedy. For example Detroit, which is near the middle of the pack in total points scored per game, actually has the league's third most efficient offense. In contrast, Charlotte, which only averages 2 fewer points per game than Detroit, actually has the league's second-worst offense.
So what does this mean for the Celtics?
One online commentator recently observed that the Celtics "haven't been a better offensive team since the trade: they're scoring at the same pace as they were before the trade." But it turns out that this isn't true. What has really happened is that the Celtics have become a more deliberate team, turning the ball over less frequently and making an extra pass more frequently.
The decrease in quick turnovers and increase in ball movement since the trade has caused each of the Celtics' possessions to take longer: Pre-trade, the average Celtics game involved 95.8 possessions. Post-trade, the Celtics take longer per possession, and thus use only 92.9 possessions per game. And yet, with almost three fewer possessions per game, the Celtics are still scoring as much as we were before the trade. As a result, the Celtics' offense, which had been near the bottom third of the league in PPP prior to the January 26 trade with Minnesota, is now scoring more points per possession than all but five other Eastern Conference teams.
The Celtics have been dramatically more efficient for four reasons:
- The stellar play of Paul Pierce, who since the trade has shot 47.8% and averaged 11.5 FT per game.
- The dead-eye shooting of Wally Szczerbiak, who since joining the Celtics has shot 37.7% from three and 88% from the line.
- The emergence of rookie Ryan Gomes, who since the trade is shooting an astonishing 56.5% from the field (on nearly 8 shots per game).
- A decrease in team turnovers from 16.2 per 48 minutes before the trade (second-worst in the league, ahead of only the Knicks) to 13.8 points per 48 minutes following the trade (equal to the league average).
Of course, if the game is slower for us, it's slower for our opponents as well. As a result, our increased efficiency means that post-trade, we've improved our scoring relative to our opponents' scoring. This is why, as the example above showed, Celtic coaches are more concerned with offensive efficiency (measured by PPP) than with total points per game.
As we all know, the game is not all offense. Since the trade, our defensive efficiency has slid slightly, down to 106.2 PPP allowed (24th in the league) from 105.0 PPP allowed pre-trade (21st in the league). But this gap may have nothing to do with the trade: two weeks after the trade we lost a key defensive cog, shot-blocking center Kendrick Perkins, to a shoulder injury that kept him out for a month. In the post-trade games in which Perkins has played, our defense has allowed 105.2 PPP, almost exactly the number of points per possession we had allowed prior to the trade.
Celtic coaches also believe that defenses take longer to adjust following a personnel change than offenses do. Offense can occasionally be played one-on-one, whereas defense requires the entire team to play together. So the fact that our defense has been no worse post-trade (when Perkins plays) is quite promising for the future of this young team, which has suddenly become one of the better offensive teams in the league.
Michael Zarren is the Celtics' Basketball Operations Analyst, responsible for assisting team decision-making via quantitative analysis.