Nylon Calculus: Defending drives and defensive correlations

Nov 9, 2016; Atlanta, GA, USA; Chicago Bulls forward Taj Gibson (22) shoots over Atlanta Hawks center Dwight Howard (8) during the first half at Philips Arena. Mandatory Credit: Dale Zanine-USA TODAY Sports
Nov 9, 2016; Atlanta, GA, USA; Chicago Bulls forward Taj Gibson (22) shoots over Atlanta Hawks center Dwight Howard (8) during the first half at Philips Arena. Mandatory Credit: Dale Zanine-USA TODAY Sports /
facebooktwitterreddit

Dribble penetration is one of the most important actions for an NBA defense to handle. Preventing it completely is impossible and so a defense must be ready to both limit it, but adapt and adjust when it does happen.

With the NBA three weeks into its schedule, I thought I’d take an early look into how teams have done when defending drives. You can find defensive drive stats for last year here. Stats for this season are below, these are just adjustments from what the NBA offers at their stats page.

As I noted on Twitter, the Golden State Warriors allowed a lot of drives last year but they complimented that with great rim protection. This year, they are still allowing a ton of drives but the rim protection is a lot worse. That has led to some bad defense.

You’ll notice that some of the best defenses such as the Atlanta Hawks, Detroit Pistons, Charlotte Hornets, Oklahoma City Thunder and San Antonio Spurs are among the worst teams in deterring drives (Opp Drive Rate). In fact, the Thunder, Hornets, Spurs and Hawks are among the five worst teams in deterring drives while they all rank in the top 8 in defensive efficiency.

Read More: The Charlotte Hornets know exactly who they are

Meanwhile, the Sacramento Kings and Phoenix Suns have been among the best teams in deterring drives but they are among the worst defenses. The Boston Celtics are another team who is among the best in deterring drives but are below average on defense.

However, that inverse relationship doesn’t hold everywhere. The Miami Heat, Utah Jazz, and Chicago Bulls rank in the top 10 in deterring drives and defensive efficiency. The Mavericks are second in the league in defending drives which has helped fuel an above average defense. No doubt adding Bogut, a great rim protector who led the league in Defensive RPM last year, was a big addition. The team has completely fallen off a cliff on defense when he sits.

You can see that preventing drives is not a straight path to a good defense. Some of the best defenses in the league don’t necessarily excel at keeping dribblers out of the middle of the floor.

You’ll notice there are a few key statistics when it comes to defending drives: Drive rate allowed and points allowed per drive. The former measures a teams ability to deter drives while the latter measures a teams ability to defend drives. Theoretically, a team wants to do one of two things when defending drives: deter drives altogether or when a player drives, make sure to protect the rim and close out on three-point shooters which will ultimately keep those drives from turning into points. So which is more important?

I looked at defensive drive statistics for last year and compared that with defensive efficiency as well as a number of other defensive statistics. We can see the correlations on a game to game basis here:

opp-drive-correlations-game-by-game-2
opp-drive-correlations-game-by-game-2 /

A few things are notable about this graph. First, with the exception of opponent 3PT% allowed, none of the correlations are that strong for defensive efficiency. Rim protection and defensive efficiency have a moderate correlation while the correlation between defensive efficiency and points allowed per drive is weak. What does this mean?

I think if anything, this shows the randomness involved in individual games. 3-point percentage is highly variable and the best way to defend 3s is to not allow them. The fact that this was the most correlated with defensive efficiency suggests there is a large amount of randomness in each individual game (because of the randomness in opponent 3PT%). And this is not surprising: sometimes, you run into a player who is unconsciously hot on a given night and there’s nothing you can do about it.

So a better way to look at these correlations may be to look at it for the entire season.

season-correlations-for-opp-drive-rate-2
season-correlations-for-opp-drive-rate-2 /

Immediately, we can see many of the correlations are stronger. It is worth noting that this was only for the 2015-16 season but still, we can see that the drive statistics have a much stronger correlation with many of the other defensive statistics.

Start with points allowed per drive (which is Opp Drive Pts% in the graph). The correlation with defensive efficiency actually remains the same and its correlation with the other defensive statistics is generally weak with the exception of rim protection (Def Rim FG% in the graph) where there is a moderate correlation. This makes sense: a team that does well defending drives is likely to also be good at protecting the rim. We also see there is almost no correlation with deterring drives (Opp Drive Rate) and defending drives (Opp Drive Pts%).

Perhaps the most interesting aspect of this graph is when we look at the correlations with the other defensive statistics for deterring drives (Opp Drive Rate). Many of them are surprisingly negative with the exception of rim deterrence, which makes sense because teams that allow a lot of drives are probably also likely to allow a lot of shots at the rim.

The most interesting element here may be the negative correlation between deterring drives and defensive efficiency. We would not normally think that teams that allow more drives are actually better defensively but there’s a moderate negative correlation. It’s likely that the better defensive teams have a strategy where they funnel drives to their rim protectors. And as we can see in the graph above, there’s a weak negative correlation between rim protection (Def Rim FG%) and deterring drives (Opp Drive Rate) indicating that teams who allow more drives are better at protecting the rim. One team that displayed those tendencies to the extreme was the aforementioned Warriors. They were dead last in deterring drives but were very good at protecting the rim.

I’s important to note that while a team may give up more drives and attempt to funnel opponents to their rim protectors, it’s vital a team still defends those drives well. There’s still a positive correlation between defending drives and defensive efficiency, which brings up an interesting idea: allowing penetration isn’t necessarily a flaw, its how you handle it which determines how good your defense is.

The strongest correlation in the graph above is deterring drives and deterring 3-point attempts. There’s a strong negative correlation between the two which indicates that teams who are allowing more drives are also deterring three-point shots at a higher rate. If teams are funneling opposing ball-handlers, which are usually guards and wings (the league leaders in drives are all guards and wings) into their rim protectors but are able to guard against the three point line, it leaves the opposing player attempting to make a play over a taller player, which is significantly harder.

Related Story: A quick look at deflections

And for teams who have great rim protectors, this leads to them being more involved in the defense (you’ll notice that the players who have defended the most shots at the rim are good rim protectors or have been in the past) which is definitely a good thing.