Paint Protection Starts With Transition Defense

http://nyloncalculus.com/2014/09/22/paint-protection-starts-transition-defense/



Paint Protection Starts with Transition Defense
Posted on September 22, 2014 by Andrew Johnson



I’m pretty sure it’s in the NBA orientation material for new coaches to pledge to play an ‘uptempo’ style of basketball, since virtually every new hire starts out stating this intention. Once the games start, however, results often vary.

In addition to whatever hand-out material the new coaches get from the league office, there is good reason for coaches to preach pushing the ball. Whether it’s casual observation of the games, any stats analysis, or applied logic—all tell us that it’s easier to score before the opposing defense gets set. The primary driver of that is the increased ability to get the basket for an uncontested dunk or lay up, with the open trailing three a distance second choice and the dreaded pull up jumper in transition, or PUJIT, a deserved subject of derision.

Jeff Haley, who runs the college basketball stats site, Hoop Math, used play by play data to breakdown the difference in shot selection based on the time left on the clock (removing shots after an offensive rebounds) in the NCAA. As one would expect shots in transition are much more likely to be right near the basket, in fact 70% of all shot attempts in the first 5 seconds after a steal were at the rim. And the share of shots at the rim declined as the shot clock wore on, no matter how the possession started, as you can see in the charts below, courtesy of Haley and Team Rankings.




Under the heading of ‘Miscellaneous,’ the NBA tracks both Point in the Paint (PITP) for both the team and their opponents and fast break points (FBP) by team and their opponents. Unhelpfully, FBP is defined by the NBA data glossary as simply ‘fast break points,’ so we’re a little in the dark as to how exactly it was calculated. But I will give the NBA the benefit of the doubt and put off data cleaning seven years of play by play data myself; at least for the purposes of this post so I can use their data.

Between the 2007-2008 season and last year, teams gave up an average of 13.2 Opponent FBP and 41.2 Opponent PITP. The best season stopping opponent fast breaks during the study period was by the 2007-2008 ‘Big Three’ lead Boston Celtics. That team also had the lowest number of opponents PITP at 34.1 per game. The worst team giving up opponent fast break points was the 2010-2011 Golden State Warriors, who gave up a lot of opponents point in the paint, but not quite as many as the 2013-2014 Lakers did, surrendering 49.2 points in the paint per game.

As expected the analysis shows a strong positive correlation between points in the paint and allowing fast break points, I added fast break points scored by the team as a partially control for pace (it doesn’t quite make sense to completely control for pace, since that measure is symmetric with offense and defense, and therefore includes the main variable we’re testing for: opponent’s fast breaks).

Those two variables give us a R2 of 0.46, meaning we can explain approximately 46% of the variation in teams points given up in the paint with a model equation of:

Opponent PITP per Game =  24.39 + .18 * FBP + 1.086 * Opponent FBP

In other words, we can add a little over 1.08 points to our expectations of point in the paint surrendered for every Opponent FBP.

Now the interesting part to me is that we can apply this formula to come up with an Expected Opponent Points in the Paint based on the team’s transition defense with some control for offensive pace. Then in the final step I looked at which teams’ defense did better or worse protecting the paint to give us a measure of the teams’ half court rim protection.

Here are the data for last year:




Key take aways from the table above:

Based on this data, Steve Clifford did an excellent job in Charlotte last year establishing a great transition defense, the best last year and among the best in the last seven years. This wasn’t an accident, read Zach Lowe’s mid-season breakdown including input from Clifford.

The Lakers had terrible defenses in both transition or half court, just really terrible.

Philadelphia had a terrible transition defense, due in part to all of the turnovers, but weren’t terrible defending the rim in the half-court.
Minnesota looks similar to Philadelphia, their half-court rim protection doesn’t come back that bad overall, but their transition defense was horrendous
Most of the teams who are in the top or bottom in half-court rim protection are exactly who we would think, with the Indiana Pacers at 3.14 points fewer per game allowed in the paint than we would expect followed by the Chicago Bulls and Memphis Grizzlies

One of the most interesting results is the Clipper’s poor result in half-court rim protection, in contrast to the trio above with three legitimate DPOY candidates. The Clippers tie with the Portland Trailblazers at the bottom of the list.

The Boston Celtics ‘rim protection’ problems look to be real as well and the Toronto Ghost Defenders might have some work to do.

The primary point of this exercise has been to get a bit of separation of the half court defense of the basket that I think most of us, certainly including me, think of when we bring up ‘rim protection’ from transition defense, which can be thought of as a subset of protecting the rim. And, as we can see not all teams are equally adept at both.



bob
MY NOTE:  Making it harder for opponents to get you back on your heels and ram it down your throat helps improve your defensive stats.  In other news, water is wet.  The purpose of the column was to talk about the need for transition defense, great, but after my eyes stopped spinning like slot machine in Vegas I realized there was another take-away from this that wasn't mentioned:  the value of generating steals and value of minimizing active turnovers.  The first chart is stark in its meaning.  When the offense was started with a steal the percentage of shots at the rim (hit or miss) was at its highest, by far, at 70% with 5 seconds or less taken off the clock, and when the fga was taken with more than 7 seconds gone the percentage dropped to 40%.  As the time dragged out into more seconds, the rate of success declined.  Why?  Because the defense got back and set up.  If the key to success is transition defense, then the key to success is also a team's ability to create transition offense and force the other team to execute their transition defense well (or else).  The second two graphs were largely meaningless, as far as I'm concerned (of course the percentage of total fgas coming off of steals is a small %.  duh) and, to be completely honest, I don't understand the big chart at all.  "Expected Opponent's Points In The Paint"?  Huh?  If they were in the paint, they were in the paint. You don't need to expect them, just count them.  If Philly is looking good, per that chart, then that chart is garbage because Philly didn't play transition defense last year.  Oh, they ran like Stan Albeck was lashing them with a bullwhip, but they didn't get back.  Where they did very well is shooting the passing lanes and making steals in the half court and, with their "all-offense-all-the-time" game, that turned into a lot of fast breaks and that shows up in their FBP number.

So, the question I'm left with is "how to create more steals?".  Rondo does his gamble defense thing and I hate it.  I am an advocate of moving your muffins on defense and not just hoping you get lucky.  Bradley's defense was based upon staying in front of the man and darting in to try and get a finger on the ball.  I predict Marcus Smart will be the same, with that insane 6'9"+ wingspan.  A tip, someone picks up the ball and everybody's running and we're off the to races.

Shooting at the rim improves fg% and shooting at the rim early in the shot clock improves it significantly. What a concept. Everything old is new again.



.

What does Andrew Johnson know about basketball?  He's a 200-year-old ex-President of the Country who arose to the position because his boss got assassinated.  No sports credentials at all.

Yes, good transition defense can make things easier for your interior defenders.  Yes, shooting early in the shot clock makes it more difficult for the defense to get set.  Ho hum.  Could we possibly lobby to start training camp a week early?

Sam