First Quarter Check-In: The Defense

As I sit down to write this, the Trail Blazers are an impressive 15-4, good for third place in the Western Conference. With 19 games in the books we are about a quarter of the way through the season and have enough data to start checking in on teams. Caveats of "small sample size" will slowly begin to fade away as "the way teams are playing" becomes "the way teams are".

So how are the Blazers doing? It's a simple question, but actually a pretty large one once you get into it. For that reason, I'll be giving you an update on the defense and the inimitable Evans Clinchy will follow up with his perspective on the offense this Friday. So stay tuned.

*    *    *

So how is the Blazers' defense doing?

The short answer is great! The Blazers currently have a defensive rating of 101.4, good for fourth in the league. That's improved from 107.4 and 16th in the league last year. The Blazers haven't just reached their goal of a top ten defense, they've pushed their way into the VIP room, rubbin' elbows with the Thunder, Grizzlies, and Spurs. At the end of the day, that's really the only number that matters, but stopping there is a little unsatisfying because it doesn't tell us how they've improved and if it's sustainable.

But before we jump into all that, let's take a step back. Last year, the Blazers got out of the cellar and into the middle of the pack by changing their defensive scheme. They began dropping their bigs back on the pick and roll, keeping them closer to the hoop and taking away the three point line. The Blazers wanted the two people involved in the pick and roll to take the shot rather than help and be forced to scramble as opposing teams passed to open shooters. They executed these principles effectively, allowing the fewest three pointers and assists in the league.

The weaknesses of this style of defense were also clear. Since the Blazers' guards struggled to get through screens, opposing ball handlers had lots of space to operate. They used this space to shoot open jumpers and, more importantly, drive to the basket. Over the summer, coach Terry Stotts decided to double down on this defensive philosophy but tweak it slightly in the hopes of allowing fewer shots at the rim. This meant improving in transition and being more aware and active on the weak side.

To his credit, the season has pretty much gone according to plan. The team hasn't drastically changed, but they've made modest improvements in a number of areas including the ones Stotts prioritized. We can start to see these improvements by looking at the Blazers' four factors.

The four factors are opponents' effective field goal percentage (Opp EFG%), opponents' turnover percentage (Opp TOV%), defensive rebounding percentage (DRB%), and free throw rate (FT/FGA). Statisticians have found that these four statistics are the best predictors of a good defense and their relevance makes intuitive sense.

On any given possession, the defense is first trying to prevent a shot altogether, then force the opponent to miss if they do get a shot off, and finally rebound the ball to end the possession. The first three factors measure how often a defense does each of these things. Opp TOV% measures how often a defense forces a turnover preventing the opposing team from even attempting a shot. Opp EFG% adjusts traditional field goal percentage taking into account that three point shots are worth more. This essentially captures how many points the opposing team gets on the shots they do take. Finally, DRB% measures how often teams get the rebound ending the possession and preventing a second shot. Free throw rate then rounds out the group capturing how often opposing teams get to the charity stripe.

The Blazers have fouled and forced turnovers at roughly the same rate but they're rebounding a bit better and are now elite at forcing misses. We don't need to talk about low turnovers and avoiding fouls much since it's just a product of the Blazers' scheme. I was hoping improved weak side positioning would increase the turnovers a little but that hasn't happened.

The jump in rebounding requires a simple, two word explanation - Air Sasquatch. Chris Kaman is tied with Tim Duncan for the eighth best defensive rebounding percentage in the league. Not only has he been a beast individually, but his presence has allowed Joel Freeland to shift to PF and the Blazers haven't played a single minute of small ball. By having two seven footers on the floor at all times and relying on the skills of LaMarcus Aldridge and Kaman the Blazers are quickly becoming one the best defensive rebounding teams in the league.

We should also give a shoutout to our point guards. This year Damian Lillard and Steve Blake are pulling down two extra defensive rebounds a game compared to last year's combo of Dame and Mo Williams. Part of this is new personnel and their individual effort, but Stotts' emphasis on weak side positioning is also putting them in a better position to rebound.

*Apologies for the video quality on some of the clips

In this clip, Lillard is doing what's called "nailing" (insert middle school joke of choice here). On nearly every basketball court there's a nail marking the center of the foul line and teams use it to mark where weak side defenders should be. The Blazers are doing a much better job of getting to the nail rather than staying a foot or two farther out along the perimeter. This slight change of positioning helped Lillard get to that rebound before anyone else. Barring injuries, there's no reason to believe the Blazers won't maintain their improved rebounding rate over the entire season.

The same can't be said for Opp EFG% unfortunately. There are three possible explanations for why the Blazers' are allowing a lower percentage this year. First, that the defense is forcing opponents to take tougher shots. Second, that the defense is forcing opponents to miss the same shots by contesting better. Third, that opponents are just shooting poorly. The first two might be sustainable over an entire season. The last one, not so much.

Let's take a look at them in order. Below is chart showing what percentage of opponents' field goal attempts came from what distance and it helps us understand if the Blazers are forcing teams to take tougher shots.

This defensive possession was so great that I stood up and cheered while everyone else looked at me like I was crazy. Damian starts by completing avoiding a Jon Leuer screen. This lets Kaman stay close to Leuer, preventing the jump shot. Kaman and Batum totally snuff out a back door and side pick and roll. Lillard stunts to prevent another Leuer jumper, and then Barton smartly sags off of Allen to contain Udrih. It ends with Udrih trying to get to the rim through a thicket of arms and losing the ball out of bounds. Just excellent.

Notice how Allen Crabbe has sagged closer to the paint and then breaks towards the hoop immediately when Kaman leaves his man to help. This is called "pinching" and it only works if you do it that decisively. Last year, the Blazers were a often a step late giving up easy layups in those types of situations.

This clip highlights another piece of evidence that the Blazers are more active on the weak side. Last year, Blake wouldn't have sagged that low to help on Afflalo. This set up a chain of passes with Blake and Batum moving in concert to prevent the drive and contest the three. If Chandler makes that shot it's an assist. So far the Blazers are still top five in preventing assists, but they no longer lead the league by a healthy margin. Plays like these are a big reason why.

Outside of the half-court, transition defense also plays a major role in preventing three pointers and shots at the rim. The Blazers are playing at roughly the same pace as last year while allowing 1.3 fewer transition points per game, according to teamrankings.com. They're still right around average, and I would guess Stotts has been a little disappointed that the team hasn't made more of an improvement in this area.

Note that we're talking about very small changes here. Since the Blazers allow roughly 85 attempts per game, forcing 52% instead of 49% in the midrange constitutes moving less than three shots a game. The Blazers' focus and improved positioning is certainly sustainable over the long haul, but the line between success and failure is razor thin. The Blazers will need to keep up the defensive intensity and focus if they want to hold on to this modest improvement.

Unfortunately, even if they do it's going to be nearly impossible to prevent some regression. Blazers' opponents have shot quite poorly so far, especially from behind the arc.

Thirty-five percent for long twos and twenty-nine percent from three are not just a little bit low, they're obscenely low. The lowest percentages allowed last year were 36.6% and 33.2% respectively and this is the first piece of evidence that the Blazers' hot start might not be sustainable.

Unfortunately, this is not a small problem. If you assume the current shot distribution remains constant but the midrange and three point shooting percentages return to what they were last year, our defensive rating would increase by about 5.5 points. That would take us out of the top ten and back to the middle of the pack.

That projection isn't really fair because everything is constantly changing. Opposing teams are shooting better in the 3 -16 ft range this year and those percentages might fall. Not to mention that the Blazers are allowing the fewest fouls shots in the league over the past 10 games after an uncharacteristic, hacky start. Either of these trends or others could offset some of the damage as opposing teams start to shoot better. However, those counterbalancing improvements are likely to be modest in comparison and the Blazers probably can't afford to let those percentages climb back to what they were last year. They need to be contesting more shots in addition to all of the other improvements I've already discussed.

Lucky for us, the new SportVU tracking cameras give us some pretty cool data to look at questions like this. For those that aren't familiar, the SportVU cameras track the location of the each player and the ball at all times. This gives us not only the precise location of each shot but also the location of the closest defender. As a result, we can measure how many shots the Blazers contest, defining a shot as contested if there's a defender within four feet of the shooter.

Here's the data for shots between 16 feet and the three point line against the Blazers. All the shots are tracked by stats.nba.com and then made searchable by nbasavant.com.

Side Note: If you haven't seen NBA Savant's Shot Tracker do yourself a favor and go play around for a bit. See which Lopez brother has the better hook shot or who's gotten the most uncontested three point shots in the league. Hint: the answer to both wears red and black on weekends.

Intuitively, the more shots you contest, the lower percentage your opponents should shoot and this relationship has been seen across the league as a whole. However, it's unclear how reliable that relationship is on a team by team basis because it misses how a defense may systematically pick who to leave open. For example, a great defense may choose to give Draymond Green an uncontested jumper in order take away a Stephen Curry shot all together. They might contest a lower percentage of shots but still give up fewer points because they never let Curry get a shot off.

We only have one year of data but this seems to be an important consideration. Only three teams were top ten in both the percentage of midrange shots they contested and in FG% allowed. In fact, Detroit ranked eighth in the percentage of midrange jumpers they contested and dead last in the FG% allowed. Conversely, the Thunder contested the lowest percentage but were sixth in FG% allowed. So perhaps we shouldn't read too much into the Blazers jump from 18th to 5th best at contesting midrange jumpers. However, given that the Blazers' coach, personnel, and defensive philosophy have largely stayed the same these complicating factors are less relevant. As a result, I would expect the Blazers to give up a lower field goal percentage this year now that they're contesting more shots.

Unfortunately, that same logic leads me to believe the Blazers' uncontested field goal percentage allowed is unsustainable. It's not like the Blazers have drastically changed which opposing players they're leaving open. There's obviously some degree of randomness so maybe the Blazers were unlucky last year to give up the third highest uncontested FG%. But the lowest uncontested field goal percentage allowed last year was 38.1% so there's no way the Blazers can maintain their 34.2% over the entire season.

Put those two together and teams are gonna start shooting better but probably not as well as they did last year. This is a testament to the work Lillard has put in over the summer along with the rest of our guards. I wrote a few weeks ago how CJ and Lillard were doing a better job of getting over screens and I think both have moved from "consistently bad" to "inconsistent". They still make too many mistakes and Lillard was especially bad against the Clippers and Nuggets. But those games are becoming outliers rather than the norm and getting over screens has helped them stay close as opposing guards pull up for jumpers.

Now they're not going to block shots very often but they have been bothering opponents shots more consistently, and this explains why the Blazers have been able to contest more shots in the  midrange. If they keep improving over the next few years the Blazers' backcourt could become a defensive strength rather than a glaring weakness.

While defending the midrange was a struggle for last year's crew, the three point line was certainly an area they owned. So far this year, the Blazers appear to have built on that success.

Side Note #2: The Timberwolves and 76ers are only contesting 9% and 10% of their opponents' three point attempts respectively. Ouch.

Not only are the Blazers allowing fewer threes a game, they're also contesting more of the triples opposing teams do get off. But these improvements have been slight and there's no way they're enough to explain why other teams are shooting just 31% on uncontested threes. In the end, the conclusion is very similar to the one we had about midrange shots. Opponents' shooting percentage from three will certainly rise but probably not as high as last year.

The key question is where these percentages will settle in between where they are now and where they were last year. That, along with the Blazers' ability to maintain their other improvements, will determine how good the Blazers' defense will be at the end of the season. There's no doubt the Blazers are better, but it's going to be difficult to maintain their top five status as those shooting percentages regress to the mean.

We're a quarter of the way through the season and it would be hard to ask for a better start. Rather than a single revolution, the Blazers have improved their defense in a lot of small ways, and we see that in both the film and the stats. The difference between good and great is always minuscule and everyone involved needs to emphasize the little things if they want to keep up with an insane Western Conference. Even if they do that, the Blazers' defense is probably at its highest point for the season, artificially inflated by their opponents' poor shooting. I will be surprised if they can maintain their top five ranking for the entire season, but for now let's enjoy all the clanging and the wins that go with it.