Play The Shooters

Near the half way juncture for the season, one trend has become clear: The Pelicans are the their best when they play shooters

Over the past few weeks, I have had a thought rattling around in my brain and was curious if there was any empirical data to support it. The hypothesis was the Pelicans are better when they play two or more shooters together, as opposed to when they play just one. I am defining “shooter” on the Pelicans roster as one of four players: CJ McCollum, Trey Murphy, Jordan Hawkins, and Matt Ryan. These are the only four players on the Pelicans roster I believe have the shot credibility to warp defenses.

Below is a chart that illustrates the Pelicans’ net rating every different combination of those four players that has gotten minutes together juxtaposed with lineups that feature exactly one of them. These are garbage time removed possessions with Liddell, JRE, Seabron, and Gates all off the court.

As expected - the multi-shooter lineups thoroughly outclass the single shooter lineups. The only single shooter lineup that is positive is the one that features CJ, and he spends more time with the Pelicans’ premier players than any other shooter. Meanwhile, the lowest multi-shooter configuration comes in at a +9.5 per 100 possessions net rating.

Let’s dig in a little further. Because the number of possessions vary wildly, I did a possession weighted average for the multi-shooter and single shooter lineups. Here were the results:

There are a few things that jump out to me immediately. Firstly, single shooter configurations have played over 300 more possessions, or roughly three games worth, than the multi-shooter configurations. The actual number is much higher. You see, the multi-shooter configurations contain overlap. For example, the CJ + Trey + Hawk combo by default includes the Trey and Hawk subset within it’s minutes. When you remove all the overlap, the actual amount of on court possessions the Pelicans have used a multi-shooter configuration is 987 possessions. This means the Pelicans have been in single shooter configurations almost twice as much as multi-shooter configurations.

After you digest that bit of information, you remember there are several “no shooter” configurations that we are not even counting here. But more on that later.

Secondly, the difference on both ends of the floors is stark. The multi-shooter lineups are not only elite on the offensive end, they are significantly better on the defensive end than the single shooter lineups. This runs counterintuitive to how the coaching staff has constructed lineups. I think most people would agree that the staff considers CJ, Trey, Hawkins, and Ryan to be among the weaker group of defenders on the team. Yet playing two or more of these players together has produced significantly better defensive results than playing just one. Furthermore, the worst defensive shooter combination of CJ and Hawk, sported a better defensive rating (116.5) than the best single shooter configuration of only CJ (116.7). Whatever they are seeing on film, simply is not bearing out by the numbers.

What About the Zero Shooter Lineups?

There is one super fascinating wrench in the data, and that is the NO SHOOTER lineups. These lineups of course are the ones without any of CJ, Trey, Hawk, and Matt on the court. In 459 no shooter possessions (also minus Liddell, Gates, Seabron, and JRE to remove garbage time), the Pelicans are a +18.28. The bulk of these possessions, however, feature all three of the Zion, BI, and JV trio, and almost all of them feature at least two of the three.

The Pelicans had a ton of success running the Dyson, Herb, BI, Zion, JV starting lineup and that success is carrying a lot of this data. However, the no-shooter configurations are shooting 39% from three, with even the above starting 5 clocking in at 40% from deep. The combination of unsustainable shooting, defense generated transition play, and offensive rebounds, the Pelicans found a way to make no shooter groups work.

Concluding Thoughts

Over the past there seasons, I have frequently felt that the Pelicans’ offensive process leaves a lot of change on the table. The roster has finally been bolstered by multiple gravity shooters, yet the team almost goes out of their way to avoid playing them together. For years fans and analysts alike have wondered how Brandon Ingram and Zion Williamson would operate in space and it appears the team finally has the opportunity to experiment here.

The bulk of the team’s lineup configurations are single shooter oriented. This minimizes huge advantages offered by the roster construction. On balance, those groups get outscored and perform poorly on both ends of the court. And it makes sense - a group that features only one shooting threat is easy to defend. If the talents of Zion Williamson and Brandon Ingram are not on the court together, the single shooter group has virtually no shot of surviving.

Moreover, the no shooter groups shot unreasonably well when they are on the court. It is irresponsible to project lineups that feature both Dyson and Herb along side Zion, BI, and JV to shoot 40% from three. Teams will play the percentages and over a long enough time, the Pelicans will lose to the math. Zion and BI in theory should always be able to carry mediocre lineups due to sheer talent, but that’s kind of my issue. Why is the system not designed to make their jobs easier?

Every step of the way, both the eye test and the data has strongly supported playing multiple shooters, and every step of the way the Pelicans have been obstinate about accepting those configurations. The power lineup, you know #thatlineup, of CJ, Hawkins, Ryan, Defender, Jonas was dominant to start the season, and their play became undeniable even to the coaching staff. Ryan may have gotten hurt, but Murphy has been available to try and replicate the same. In the 16 games Murphy has played, the Pelicans have attempted the CJ, Hawkins, Murphy, Jonas foursome for just 19 minutes. They outscored the opponent by 19 in those 19 minutes.

For the third season running, the Pelicans just do not prioritize shooting. Make no mistake, this is a top down mentality. The coaches set unattainable goals of shooting 40 threes per game, while one of their franchise players has decreased his attempts each month of the season (Ingram is taking 2.1 threes per game in January). Somehow the team is supposed to hoist threes, but the coaches won’t play lineups with more than one shooter. For a team that emphasizes sharing the ball, shooting attempts become a simple probability game. With one shooter on the court, every time BI or Zion kick the ball out, there is only a 25% chance the recipient will be a shooter. When two shooters are on the court, the chances double, as now two of four passing options are shooters. The number triples to a 75% chance of passing to a shooter when three are on the court.

In my opinion, they should design a rotation that ensures two shooters are on the court at all times. Maybe it is impossible to construct such a rotation, but at the very least there needs to be an effort to reverse the current 2:1 ratio of single shooter lineups to multi-shooter lineups. The data is incredibly in their favor, I’d love to see a meaningful attempt to explore this space further.

Look, I understand that the team is winning right now. They are a very good team, that I think would be good regardless of the style or system they play. I personally feel there is a range of outcomes a team like this can have, with the vast majority of them falling under “good”. Their goal, should not be settling for good, but rather pursuing outcomes beyond the standard deviation. As David Griffin mentioned in a broadcast interview lately - the Pelicans had the best net rating in the league since the IST, but they did not have the most amount of wins. I think there’s change left on the table that is easily cleaned up.