Breaking down James Harden's "new and improved" defense

There has been a pervasive narrative this season perpetuated by talking heads and Internet discussion boards, all surrounding James Harden's defense. The story goes that he was bad at defense and then last year, he saw a 10-minute Youtube clip of him getting embarrassed defensively.

We're told he felt ashamed by the Internet and fans so he worked hard over the summer and now he isn't so bad at defense and the Rockets in general are a much better defensive team because of it.

It's a good story. It may even be contributing to his front-running campaign for MVP at the halfway mark. The problem is that statistically, there is not a lot of basis for its truth.

So, what do we know?

Thanks to the player tracking data over at NBA.com/stats we can get a pretty good idea about how an individual player has performed defensively over the course of the last couple of seasons. (The SportsVU data only goes back two seasons.)

One of the most helpful statistics in understanding a player's defensive contribution is known as Diff%. This stat measures the impact a defensive player has on individual shooters by quantifying the increase or decrease in field-goal percentage shot by an offensive player when guarded by the defender. So a good Diff% would be negative, because when a good defender is guarding a shooter, the shooter's fg% should theoretically dip below what they shoot on average.

For instance, Anthony Davis has a Diff% of -8.9 against 3-point shooters. That means that when being guarded by Anthony Davis, a player will shoot about 8.9% worse from the 3-point line than they normally would. This aligns with our understanding that Davis is a defensive beast and that you shouldn't let him contest your shots because he'll punk you in a very public fashion.

So how does 2014-15 James Harden stack up against 2013-14 James Harden? This year his overall Diff% is -2.3. Last year, The Beard finished the season with a Diff% of -2.4.  This does NOT align with the narrative.

"So you're saying he's worse at Defense so far this season than he was last year?"

Well no, not exactly that, either.

Things start to get wonky when you look at Harden's Diff% against specific distances from the hoop.

Last year Harden's best defense came against 2-point shots from further than 15 feet out in which he posted a Diff% of -1.2. Against 3's he posted the devastatingly average Diff% of -0.2. His inside defense was understandably lackluster considering his weight (225 lbs) and his height (6'5"). On shots closer than 6 feet, his Diff% was 4.9.

This season his Diff% against 3-point shots is a remarkable -5.8. Against shots greater than 15 feet? -4.6.

His overall Diff% is worse this year because his defense in the paint has dropped off so dramatically. He's allowing shooters within 6 feet of the hoop a 7.1 percent increase in their average and everything within 10 feet gets a 4.1 percent bump.

To say The Beard has had some kind of defensive renaissance would be misleading. He's very clearly making a concerted effort to defend better on the perimeter, though it would appear that on the low-block he's become a significantly worse defender than he was last year.

Explanations for this will vary from the decrease in Houston's center-depth in a post-Asik world to Harden getting easier perimeter assignments with the addition of lengthier wing-defender Trevor Ariza. It could even be that Harden only has so much energy to divide amongst his efforts. I however, am a firm-believer in Occam's razor. It seems to me that Houston just isn't a very good defensive team around the hoop and that they're damn good at defending the arc.

The overall team defense seems to lend itself to the kind of numbers that Harden is producing. When defending long 2's and 3-point shots, Houston is second in the league behind Portland but Houston's defense slides further and further down the list the closer you get to the basket. Opponents shoot 62.6 percent against Houston from the restricted area. That puts the Rockets near the bottom of the pile at 24th against inside shots whereas last year, they ranked closer to the middle of the pack at 15th.

On a team with an injured Terrence Jones and a Dwight Howard that missed 8 games from November to December, it's understandable that the team's overall post-defense is going to take a bit of a dive. It's also understandable that Harden's numbers would be reflective of that.

Don't let the narrative fool you. James Harden was never, from a purely statistical standpoint, THAT atrocious on defense. As Daryl Morey has stated, anyone's worst 10 minutes of work from the past year would make for an embarrassing Youtube clip. Yes, this year we've definitely seen an improvement in his perimeter defense. But, to go along with that, there's been a drop off on his efforts inside. It's almost zero-sum.

What should concern you, as a fan, is his contribution to the Rockets ability to win. From that perspective his defensive contributions have been almost static.