Oklahoma City Thunder film room: fast-break defense keeps OKC competitive

Through the season’s first half, much discussion has centered around the Thunder’s need to generate fast-break points to supplement the team’s half-court offense. Such sentiment is doubly understandable following Oklahoma City’s offseason loss of a certain efficient volume-scoring former MVP. Further, a staunch transition attack should take precedent for any team with which the uberly athletic Russell Westbrook sprints end-lines.

However, for the Thunder, a prominent trend has arisen this season concerning another key element of fast-break execution. Surprisingly, the Oklahoma City Thunder has become the second-best team at preventing transition points, a startling fact when you consider the team’s youth and lack of roster continuity. In two separate games this season, the Thunder have allowed zero fast-break points: one of which was against the Phoenix Suns —whom currently rank second in the league in fast-break points per 100 possessions at 18.1.

How have the Thunder, whom were mediocre in transition defense last season, turned things around so dramatically?

To reach a reasonable conclusion, there are several steps that we will take in this article. The end goal is to identify the overarching methods used to deter transition scoring, but there are sub-goals that will give a better understanding of both the generic facts of fast-break scoring and the team approach to defending on the break.

1. Determine Statistical Correlations: Using league data to determine what might improve or diminish transition defensive effectiveness will allow us to locate the Thunder’s stead in relation to the league’s average, perhaps giving insight into why OKC has been so effective.
2. Examine Personnel: Different players could be the sole contributors to the improved defense, whether based upon skill set, positioning, or game plan. Identifying the players who strengthen or weaken the transition defense could present reasons for the improvements.
3. Off-season Changes: The Thunder roster saw major overturn during the off-season. Could this have had an effect, and how significant of an effect would that be?
4. System/Structure: Sometimes a small change in approach can have a radical effect on the end results. What is the Thunder doing to prevent teams from scoring on the break?
5. Film Study: Using cases from the Clippers game in which the opponent scored 0 fast break points, visual examples of why OKC has been successful will be used to summate the findings from previous sections.

Statistical Correlations

To determine factors that improve or diminish the ability to defend in transition, other statistics have to be examined in relation to fast-break points. This can then be used with the data from OKC to determine if it helps explain the Thunder’s success or if OKC is an outlier.

Prior to presenting the data, I’d like to preface it with a few disclaimers.

Disclaimer 1: The correlations presented are from one season of data. This means that there are only 30 data points, and with so few points, the end results could actually be dramatically different than the true correlations due to variance. Why not use more seasons and get better results, you ask? That leads us to...

Disclaimer 2: All of the data used to determine statistical relations was manually input into a spreadsheet. This means that it is reasonable that some of the data has been input incorrectly (hopefully with an error of .3 or less, the common error when using a numpad). However, for the sake of this article, we will assume I’m a perfect typist.

Disclaimer 3: I’m by no means a stats expert. Much of what I’ve learned has been from articles and comments. While I do have an extensive mathematical background (and will be taking a statistical analysis class this semester) I do not currently possess that expertise. The connections I will be drawing are simple, and I’ve attempted to throw out inconclusive data, but I cannot speak with the certainty of an expert.

Disclaimer 4: The NBA stats page was having some issues when I was using it. The minutes data on some of the personnel and lineup pages was clearly incorrect. However, I believe I sorted out the error and it should not have extended to the numbers I used. With that said, it is possible that the site itself has errors that were not so obvious, and I cannot account for those.

Now that you’ve had time to get ample grains of salt with which to take this analysis, let’s begin.

Offensive Rebounding

The general belief among basketball pundits is that offensive rebounding and transition defense are negatively correlated, meaning that good offensive rebounding teams will be bad transition defenders. This is generally accepted, with many coaches choosing to eschew offensive boards in favor of having players get back to the defensive end of the floor.

My hypothesis was the opposite. I believed that having greater success on the offensive boards would result in less transition baskets because teams would dedicate more players to boxing out and defensive rebounding. Empirically, this makes sense for the Thunder, as OKC is excellent at offensive rebounding and at transition defense.

The actual data, however, shows little correlation between the two statistics this season, showing only a loose, positive correlation.

From reasoning, it appears that the two possible outcomes actually tend to cancel each other out. In general, the best place to be on this graph is low and to the right. Oklahoma City is shown by the orange dot on the graph, a favorable place to be. This does show that OKC is one of the best teams at grabbing offensive rebounds while not giving up easy transition looks. The reason for this will be discussed more deeply in the structure and film sections.

Turnovers

It is fairly logical to assume that turnovers will have a positive correlation with fast break points. Turnovers, particularly live ball turnovers, gives the fast-breaking unit an opportunity to score against minimal defensive resistance. At best, the defending transition team gets favorable matchups. At worst, they allow an open dunk.

As expected, the data shows that turnovers have a strong, positive correlation with opponent fast-break points. Additionally, opponent points off turnovers also show a strong positive correlation. In fact, of all the correlations examined, turnovers proved to be the best predictor of opponent fast-break points.

This chart is initially difficult to read, but it shows something interesting while pinpointing each team’s location on the chart. The x-axis shows the turnovers per 100 possessions, with the y-axis showing the opponent fast-break points per 100 possessions. Again, the positive correlation that is expected appears. The colors highlight the opponent points off of turnovers per turnover, with dark being a higher value, and light being lower.

The first expectation for the Thunder would be that, due to a high turnover rate, teams would score more fast-break points. However, this isn’t the case, and this chart demonstrates why. Oklahoma City is located in the center of the chart at the bottom. Notice that there is no colored circle around it. This means that OKC is the best team at the league at defending after turnovers.

Digging even deeper into this stat, it becomes even more impressive. In fact, I don’t think there is a way to give this full justice, because it’s so unbelievable. But I’ll do my best.

The league average for points off turnovers per turnover is 1.137. As expected, this is much worse than even the worst defense in the league cedes overall (Portland at 1.106 opponent points per possession). OKC gives up a league-best 1.041 points off turnovers per turnover. When adjusted in comparison to overall defense, that mark would be good for 12th best in the league. Let me say that again.

If the Thunder gave up a turnover every single possession, including live ball turnovers leading to transition attempts, they would still have the 12th best defense in the league.

For this very reason, the Thunder is an outlier on the chart. Turnovers don’t hold the same correlation to fast-break points for this team as they do for the rest of the league. So OKC is somehow defending exceptionally well after turnovers. Why is that? An attempt to explain will be made in later sections.

Three-Points Misses

Bricking shots from long range, particularly when shot from above the break, has frequently been associated with transition attempts. It’s accepted that the shots rebound further on average, allowing for the run-out opportunity.

However, there was no statistical correlation between missed three-point shots and fast-break points. This could be due to increased offensive rebounds from three-point shots. But this interesting data does belie the general belief.

This lack of correlation aids the Thunder who, despite being the league’s second-worst 3PM% shooting team, still attempt almost 26 threes per 100 possessions.

Other Correlations

Some other correlations can be noted that, while interesting, don’t aid us in determining how OKC has been so successful.

Teams that allow a high number of transition points tend to have high defensive ratings. This, of course, makes sense; giving up efficient points more frequently will hurt a team’s overall defensive effectiveness.

Additionally, teams that give up a lot of fast-break points also tend to give up a large volume of free-throw attempts. Again, this is expected, as transition baskets usually come against a defense that isn’t good at getting set or preventing mismatches.

Summary

The biggest thing that is demonstrated by these statistics is that turnovers, as expected, have the biggest correlation with fast-break points. However, this does not hold true for OKC because of an exceptional ability to defend following turnovers.

However, we still haven’t determined anything about HOW Oklahoma City is defending so well in transition.

Personnel

One of the keys to success in basketball is finding player combinations that complement each other, and using guys who possess a certain skill to fill gaps in lineups that are lacking in that skill. Understanding which players are effective transition defenders helps to keep viable lineups on the floor and prevent teams from swinging momentum with a few fast-break baskets.

Using on/off splits for each player, it is pretty easy to see, generally, which players are having the biggest impact on transition defense, and the results are aligned closely with the expected values.

This chart shows the difference in fast-break points when a player is on the court versus off the court. A negative number is good, meaning that the team gives up less points when a player is on the court than when they are off the court.

As expected, the top two performers are Adams and Roberson. These guys are the foundation of the defense, so it comes as no surprise that they lead the team. Close behind are Westbrook and Sabonis. While Westbrook isn’t known for getting back, the offense runs more smoothly with him on the floor, leading to oppositional fast-break opportunities. Additionally, he has picked up a habit of frustration fouling in some transition attempts rather than staying behind the play to complain, which could have a subtle effect on his numbers. Mostly though, his good numbers probably come in part from spending so much time with the starters. Abrines and Oladipo round out the players who are contributing positively.

Not surprisingly, Joffrey Lauvergne, Anthony Morrow, Semaj Christon, and Enes Kanter hold the bottom-four slots. Each are relatively nonathletic, and none of them have a good defensive reputation. The surprise comes from Jerami Grant, an extremely athletic player who is known for shot-blocking, a valuable skill in transition. However, transition defense starts on the offensive end when a shot goes up, so there could be a lack of attention to detail that reduces his effectiveness. Additionally, he spends much of his time playing with the other poor contributors, so that also will impact his numbers.

The lineup data for fast-break points was generally inconclusive. Because many of Oklahoma City’s various lineups have only played together in limited minutes, the abstracts for a relevant fast-break sample size is lacking. Additionally, the players involved in the lineups were volatile, with Adams and Roberson both appearing in some of the worst as well as the best. Not surprisingly, though, the worst three-man lineup that has played in 25+ games is Christon-Lauvergne-Kanter.

Off-Season Changes

From the 2015-2016 season to this season, there has been significant roster turnover, with much of it occurring during the off-season. Of current rotational players, only five were on this team last season: Westbrook, Roberson, Adams, Morrow, and Kanter. The rest of the rotation is filled out with players acquired during the off-season (or during the first eight games of this season). Serge Ibaka was traded for Victor Oladipo, Ersan Ilyasova (who was then traded for Jerami Grant), and the 12th pick, Domantas Sabonis. Alex Abrines and Semaj Christon were second-round picks who were signed during the off-season. Joffrey Lauvergne was acquired for two 2nd-round picks.

The difference in personal could have a direct impact on the the effectiveness of the transition defense. At first glance, it would seem strange that losing Serge Ibaka and Kevin Durant, both considered to be good defenders, would improve this, but it is possible.

Consider this: the three rotational players who had the lowest average speed last season were Kevin Durant, Serge Ibaka, and Dion Waiters, none of whom remain with the team this season. While this isn’t conclusive by itself, transition defense relies heavily upon the willingness to run back and commit to the effort required to not allow fast-break points. Conversely, Andre Roberson has had the team’s highest average speed through the last two seasons. Overall, this season, the Thunder’s average speed has increased.

Part of this could be attributed to Oklahoma City’s youth-infusion. OKC has the NBA’s second youngest roster, and when you consider the players who are actually seeing the floor, the median-age is even more pronounced. Andre Roberson, who is 25 years old, is the second-oldest rotational player (now that A-mo isn’t playing as often) behind Russell Westbrook, who is 28. The average age of the guys who play consistent minutes is 23.4 years old.

Although, due to inexperience, youth can inhibit solid defense; it can aid as well — particularly in transition. Mistakes in transition defense can be covered by athleticism and hustle. By having the energy to commit more often, the breaks can be stopped with greater frequency.

Additionally, young players have more to prove to their coach. Kevin Durant and Serge Ibaka were going to get their minutes every night regardless. They did not have that extra motivation to hustle back on each possession. Sabonis, Abrines, and Grant simply don’t have that luxury. All three have seen games with reduced —or for the latter two— no minutes. That motivation to give extra effort has a particularly pronounced effect on transition defense.

So while the team as a whole regressed from its off-season decisions, the transition defense could have improved by virtue of the rotational roster becoming significantly younger.

System and Structure

From the film that I watched (which we will get to next), from quotes, and from some shot charts, I believe that the biggest cause of the improved fast-break defense comes from a slight change in the overall structure of the offense.

Last season, the Thunder offense was structured with either one, or both of, Durant and Westbrook above the break, Roberson in the corner, Serge either on the baseline or the wing, and Adams inside. This meant that two of the best transition defenders (Serge being the third) had to run the entire length of the floor to protect the rim. Durant and Westbrook were too lackadaisical to be that initial line of defense.

The shot charts for Roberson from last season to this campaign show the difference. First, consider last season.

Notice the areas in which Dre took most of his shots. This isn’t conducive to transition defense. Now look at this season.

While his predominant areas are still in the corner and around the rim, he has already taken significantly more threes from above the break this season than he did all of last season. This allows him to be that first defender back. It’s already proven to be effective at times this season.

I included this clip because I love everything about it. It also clearly illustrates one of the areas Dre has improved in, and that is the timing of his chase down blocks. On this play, he’s coming from above the break, which allows him to be in control and set Beverly up for the rejection. Had he been in the corner, he probably wouldn’t have had the same opportunity.

This doesn’t just occur with Dre though. Besides Westbrook, it’s not uncommon for Oladipo, Grant, or Abrines to spend time above the break. All of these guys are capable at slowing down the transition game. More importantly, they are willing to commit to stopping the break, something that has been lacking in the past. Previously, Durant often has this responsibility. Last season his on/off splits actually show that OKC was worse at defending in transition with him on the floor than with him off. This personnel change could also contribute to the team’s overall improvement.

Another schematic change has been with offensive rebounding. Donovan mentioned this earlier in the season following the win in LA against the Clippers.

“We rebound the ball heavy, but I also think too we’re not sending three or four guys to the backboard,” coach Billy Donovan said. “A lot of it is based on positions on the floor. For example, we’ve got guys above the top of the key, we don’t want them running in to crash. But if we’re playing big with Enes and Steven and our offense is situated where both those guys are in around the basket, yeah, we want them to offensive rebound, because they do it well.”

This is something that will be notable in the film section. The Thunder does not crash the board from above the break. It isn’t uncommon for four of OKC’s players to be at midcourt by the time their opponent has secured a rebound. Because of this, they are able to keep players in front of them and prevent the break.

It’s also a testament to the abilities of Steven Adams and Enes Kanter. The Stache Bros. are capable of gathering offensive rebounds despite working against five guys. Because of this ability, teams are rebounding with four or five players, allowing the Thunder plenty of time to get back on defense.

Film Study

To wrap things up, I want to illustrate some of the examples from the LA Clippers game in which LAC scored zero fast-break points. In the entirety of the game, there were only three cases in which the Clippers even had a chance at a transition opportunity, and each case was well defended by OKC.

In this clip, note where the Thunder players are when the rebound is gathered: two of the five guys have already crossed mid-court, and the other three are about to. Additionally, Sabonis has already made it within the three-point line. By the time the ball reaches the half-court, OKC has picked the match-ups, gotten in a solid defensive stance, and is able to slow the pace down.

This is a long rebound off of a miss that spells trouble. When the shot goes up, all five Thunder players are within 10 feet of the baseline. Even with them beginning to retreat as soon as the shot is released, no one is back when the rebound is grabbed. J.J. Redick has the step down the floor, but notice how he is immediately swarmed. He knows that he is at an athletic disadvantage, so he tries to use a fake to get his shot, but two other defenders arrive. Adams running hard up the floor prevents that open look, forcing the ball to swing. Suddenly, the transition opportunity isn’t there.

Look at the intensity in the motions of everyone except Westbrook. They are actively looking for a man to pick up as soon as the pass is made. That attention hasn’t been as consistent in previous seasons.

This is something I’ve noticed more from Roberson this season. When he crashes the offensive glass and doesn’t get the rebound, he tends to apply “token pressure”. Essentially, he knows that he isn’t going to get a steal out of it, but by pressuring the ball, he slows down the offense. This allows the defense to reset, turning a four-on-three break to an even four-on-four break.

Notice each player’s actions. Sabonis and Dipo head back as soon as they see Adams is unable to the rebound. Abrines sprints back to cover the cutter. Steven books it back on defense once he doesn’t have a shot at the board. Roberson also takes off as soon as Felton makes the pass.

Notice again how many defenders get back on defense. By the time the ball passes the 3 point line, four Thunder defenders are across half-court. OKC controls the pace of the Clippers’ offense by forcing them to settle into the half-court attack.

A seemingly common theme among these clips is that Sabonis is one of the first defenders back in almost every case. He spends much of his time above the free-throw line, which places him in prime defensive position. Because of the emphasis on offensive rebounding from the paint only, he is usually one of the first players back in transition.

This turnover was defended as well as possible considering that the Clippers had what should have been an easy three-on-two break. However, Russ and Dre combined to force the ball out of Felton’s hands, and Dipo was closing on Rivers, meaning the ball had to go to Redick. Roberson clearly was aware of this, because the moment Felton picks up the ball, he peels off to contest the shot. He almost gets a piece of it, and the shot rims out.

This was the Clippers’ best fast break-opportunity of the night. Again though, note how fast the Thunder players get back on defense (excluding Kanter). If you want an easy explanation as to why OKC defends so well on the break and after turnovers, this clip shows it pretty clearly.

This clip demonstrates two things. The first is the importance of communication. Abrines is the first man back, but he hasn’t noticed the man leaking out yet. Dre immediately calls it out, planning to stop the ball and let Alex handle it. The second is the effort from Abrines to get this steal. It isn’t just Andre and Steven putting in work. Sabonis and Abrines are also putting in similar effort. Even Kanter hustles back on this play. That total team effort is required in transition, and that was missing last season. Roberson even said so, saying it was fully his duty to be the first man back. This season, several guys are taking that responsibility on themselves.

This picture truly illustrates the effect of Steven Adams’ rebounding. The Clippers have five players committed to snagging the board: OKC has one. This allows four guys to get back on defense before LAC even has the ball.

One surprising element to Oklahoma City’s fast-break defensive efficacy is the lack of respect shown to the team’s perimeter shooters. Several of the Thunder players are left unguarded on offense in order for the defense to pack the paint. However, this reduces the defense’s number of guys available to run in transition.

Remember good old Derek Fisher? Every time he contested a three-point shot, he would just keep running. This is fairly normal. However, if you aren’t rushing out to contest, choosing instead to double down on help defense, those opportunities are gone. All five Clippers are within 10 feet of the baseline when the rebound is snagged because they packed the paint.

To conclude, this image shows a defense that firmly understand its goals. Each player has already identified their matchup, found proper positioning, and is retreating under control with the ball in sight. This transition discipline, though not present in all of the players, is what has made the team so good in that regard this season. The players have committed to it, and it has become a potential game changer. In a season where many of the contests are coming down to the wire, this discipline is proving and will continue to prove vital to the Oklahoma City Thunder’s ultimate success.