March 28, 2012
Inconsistency
Quantifying NBA Variability

by Kevin Pelton

The Orlando Magic's two meetings with the Chicago Bulls this month showed the extremes of the team's performance. Playing in front of a national TV audience on March 8, the Magic went to Chicago and knocked off the league-leading Bulls by a 99-94 final. When the teams met a week and a half later in Orlando, Chicago played without reigning MVP Derrick Rose. Yet the Bulls still held the Magic to 59 points, the fewest allowed in franchise history, during a 26-point win.

The divergent results fit into the narrative about the Magic. Because Orlando shoots so many threes--more than one in three of the team's shot attempts come from beyond the arc, the league's highest mark--the logic goes that the Magic's performance is unusually liable to fluctuate from game to game. Or, as you're likely to hear on TV, "Live by the three, die by the three." There is evidence behind this line of thinking. At the NCAA level, College Basketball Prospectus contributor John Ezekowitz recently found that shooting more threes is associated with larger variance and less consistency.

However, when I ranked NBA teams in terms of their consistency based on the standard deviation of their scoring margin from game to games, I was surprised to find that the Magic did not rank near the top of the league. In fact, Orlando's standard deviation of 12.8 points per game is almost exactly league average (12.7). Here, instead, are the teams that have been most and least predictable:

 MOST INCONSISTENT         MOST CONSISTENT

Team            SD       Team            SD
-------------------      -------------------
Portland       16.8      New Orleans     9.8
Philadelphia   15.2      Oklahoma City   9.9
New York       14.0      Minnesota      10.0
Indiana        13.8      L.A. Lakers    10.1
Sacramento     13.6      New Jersey     11.4

In part, what we're measuring here is how often teams play blowouts. So it's no surprise that the Portland Trail Blazers, with three wins by at least 38 points and a 42-point loss on their ledger, have been the league's most Jekyll-and-Hyde outfit. The Philadelphia 76ers, too, have tended to destroy lesser competition this season. In the case of the Sacramento Kings, it's more about getting blown out while pulling off the occasional surprising upset. On the other side, Minnesota, New Orleans and Oklahoma City have all played a number of games decided down the stretch. The Thunder has gone 10-4 in games decided by five points or fewer, while the Hornets have won just five of 14 such games.

As I noted the last time I tried to quantify team consistency, looking at scoring margin is not the only perspective available to us. In fact, it may not be the best measure of consistency, to the extent that it does not take into account location or opposition. A team that wins every game by precisely five points is actually playing two very different games home against Washington and at Miami.

Fortunately, my schedule-adjusted ratings are designed to rate performance as compared to an average team in the same situation, which allows me to use them to take a look at this second form of consistency. Here are the largest and smallest standard deviations among schedule-adjusted ratings:

 MOST INCONSISTENT         MOST CONSISTENT

Team            SD       Team            SD
-------------------      -------------------
Portland       15.1      L.A. Lakers     8.2
Chicago        13.0      Oklahoma City   8.9
New York       12.7      Utah            9.9
Denver         12.6      Minnesota       9.9
Sacramento     12.5      Washington     10.4

For the most part, these rankings look similar to the last set of leaderboards. The differences are worth noting, because they get at interesting point. Take the Bulls, who have not been especially inconsistent in terms of scoring margin but leap to second in the league when we adjust for schedule. We can interpret this as indicating that Chicago's performance is more independent of schedule than average. In basketball cliche-speak, the Bulls tend to play to the level of their competition, ensuring closer games than we'd expect against lesser foes but bigger wins than their overall scoring margin would suggest against other top teams.

Ranking teams from most to least affected by schedule yields the following rankings:

  MOST AFFECTED            LEAST AFFECTED

Team           Diff      Team           Diff
-------------------      -------------------
Philadelphia    3.2      New Orleans    -0.7
Detroit         2.4      Denver         -0.3
Utah            2.3      New Jersey      0.1
Atlanta         2.2      Minnesota       0.1
Houston         1.9      Chicago         0.3

The Sixers stand out to me as the most interesting team by this measure. What this confirms is the conventional wisdom that Philly is capable of drilling lesser teams but struggles against playoff-level competition. That's worth keeping in mind heading into the postseason, especially in the context of the 76ers' impressive point differential.

At the other end of the spectrum, Chicago is the one elite team that particularly seems to coast against lesser opposition. When I last looked at these figures in 2008-09, the Los Angeles Lakers were one of two teams that actually became less predictable when accounting for schedule. That foreshadowed the eventual champions' ability to shift into a higher gear during the playoffs. No one really thinks of the Bulls as having a switch they flip on quite like the Lakers of recent vintage, but statistically they are in a similar class. By contrast, the Miami Heat (+1.3) scores as completely average in this regard.

As for Orlando, the Magic does tend to play to the level of its competition a bit more than average, with a 0.8-point difference between the inconsistency of its scoring margin and schedule-adjusted rating. For the most part, however, Orlando is fairly typical when it comes to consistency. Splitting out the Magic's performance, the team's Offensive Rating tends to fluctuate more from game to game (a standard deviation of 13.8 points per 100 possessions) than its Defensive Rating (10.1). Orlando's defense is also better when its offense is worse. Extreme outliers like the 59-point outing against Chicago aside, the Magic is really proving out another old basketball adage: When shots don't fall, buckle down on defense.

Kevin Pelton is an author of Basketball Prospectus. You can contact Kevin by clicking here or click here to see Kevin's other articles.