Dwight Howard just finished annihilating the league in February (27 and 15, 67 percent shooting!), and he has topped my MVP ballot continuously since the quarter mark of the season, when I started really digging into the awards races.
But all along, there has been one nagging thing that I either had to conveniently ignore or explain away somehow: The Magic have played better defense with Howard on the bench. How is this possible? We’re talking about the back-to-back Defensive Player of the Year — the guy regarded almost universally as the league’s most destructive defensive force.
And yet: Orlando has allowed 102.8 points per 100 possessions with Howard on the court and just 101.32 with Howard on the bench, according to Basketball Value. It’s a small difference, but still, it defies our expectations and our eyes, which witness Howard doing ridiculous things to thwart your favorite team’s pick-and-roll attack every night.
If there’s an argument people are going to make against Howard as a potential MVP, this point will come up. I can’t ignore it. But can I explain it away?
I think I can — and without being blinded by bias.
Let’s start with this: The Magic rank third in points allowed per possession, behind only Boston and Chicago. So they’re an elite defensive team overall; the slightly higher points-allowed number they’ve posted with the big fella on the court would be the envy of two dozen other teams.
Second, Howard leads the Magic in minutes played, by a long shot, having logged about 500 more minutes than any of his teammates who have been in Orlando all season. Given that, we should expect his plus/minus numbers to look very much like his team’s overall numbers.
And third, Orlando is among the most well-coached teams in the league, especially on defense, and smart, well-executed systems are just as important as individual talent in building an elite defense. Coach Stan Van Gundy has an ideal, and he demands that each of his players commit to that ideal — stay in front of your man, don’t gamble for steals, make absolutely sure to secure defensive rebounds and generally avoid aggressive double-teaming. Howard once told me that the worst sin you can commit in Van Gundy’s system is a reach-in foul, that Van Gundy will stop practice and order some form of punishment if anyone commits one.
In short: This is a system that has turned J.J. Redick into a solid defender, and we should expect the system to function well as the Magic cycle rotation players in and out.
This is all good material for the Howard-backers, but we still haven’t really addressed the problem directly. The final and most important explanation is the simple one: Those lineups without Howard that play such stingy defense are facing significantly worse lineups than the ones Howard faces. That’s my hypothesis, anyway, and the initial research (via Basketball Value’s indispensable lineup data) has begun to confirm it.
Let’s take the following lineup: Gilbert Arenas/Redick/Jason Richardson/Ryan Anderson/Brandon Bass. That group has played about 53 minutes together, making it the Magic’s 10th-most commonly used lineup. And it has held opponents to 92.7 points per 100 possessions. Insanity!
But if you look at all 32 lineups that group has faced, you notice immediately that they’re going up against backups and, in some cases, end-of-the-bench types. Of those 32 lineups, only eight include more than two guys who are regular starters for their teams. On average, they include 1.5 starters.
Compare that to the following lineup, which has played about 81 minutes together and allowed a grotesque 114.3 points per 100 possessions: Jameer Nelson/Redick/Richardson/Hedo Turkoglu/Howard. This defensively challenged group has faced 36 five-man units. Of those 36 units, 22 include at least three regular starters. And that actually undersells it, since about a half-dozen of the 14 featuring more backups than starters come from Dallas and Oklahoma City, two teams that bring some of their most productive guys off the bench.
I haven’t gone through every important Magic lineup, but I’d bet good money that if you took each of the stingy non-Howard units (most of which include the Bass/Anderson front line), you’d find the same thing: They are stifling backups.
I’d also bet you’d find something else, which may also play into this: Those backup units are more likely to play together for shorter stretches and in head-to-head matchups that last only a few possessions. Again, check the data: That Anderson/Bass group hasn’t faced any five-man unit for more than 4.9 minutes in a single stretch this season, and more than half of its stretches against particular five-man groups have lasted fewer than two minutes.
The Howard group mentioned above typically plays longer stretches, including a combined 12 minutes against the Mavs’ most common starting lineup — one of the best groups in the league.
The case isn’t closed, but I think I’m on my way to closing it. This is the thing with plus/minus numbers: It can be dangerous to toss them out there without any context.
As a side note, that’s also why various stat-heads have created adjusted plus/minus formulas, which seek to correct for the quality of an individual’s teammates and opponents for the exact reasons I’ve discussed here. And according to Basketball Value’s adjusted plus/minus system, Howard has been one of the best players in the league. Then again, the system says the same thing about Keyon Dooling.