Good ball movement generates open shots. This is Basketball 101, so fundamental to the game's primary education it may as well come down on stone tablets to every young player. It's proven: perhaps to the relief of youth coaches around the country, SportVU player-tracking shows that ball movement in the NBA does correlate with open looks. Usually.
But before we get to the outliers, first let's go about proving that ball movement leads to open shots. To measure a team's ball movement, we'll use their average seconds per touch. This is how long a player holds the ball on average before passing it on, taking a shot, or releasing the ball via some other basketball activity. So Carmelo Anthony and Chris Paul would have very high numbers, while spot-up shooters like Kyle Korver would have very low numbers. This would show up team-wide on teams where the ball tends to stick in one player's hand instead of whipping around the court. (All stats are courtesy of NBA.com and current through the March 12 games.) We'll be using this as our proxy for ball movement.
As for open shots, that's a little easier. SportVU defines a shot as open if no defender is within four feet of the shooter. We'll look at the percentage of a team's shots from greater than 10 feet that are open. Clearly, the higher the percentage the better, but in case there is any doubt, Chart 1 show's the league's shooting percentages according to defender distance.
Surprising no one, the league shoots better when defenders are farther away.
A team's percentage of open shots is a metric that only pays attention to shot attempts. It does not care about shooting percentages. When used to judge a team's offensive design, a team's lack of talent is not an excuse. In fact, poor shooting is to the team's advantage. No one is breaking his back to run Tony Wroten off the 3-point line. They would be foolish to do so. His .261 3-point percentage for the 76ers means his 3-pointers generate .783 points per shot or the equivalent of sending a 39 percent free throw shooter to the line for two. 39 percent from the line is DeAndre Jordan-bad, and teams are now intentionally fouling him. You could say that letting Wroten fire away from deep is the more fan-friendly version of hack-a-Jordan.
So there's some noise on that end of the spectrum, and some on the other as well. There is no need to guard the weak shooters, but draw the assignment of guarding Atlanta's Korver and you better treat your coach to a Tony Award-winning production of Inseparable, a heartbreaking tale about Kyle, a man who dreams of playing a basketball game free of his conjoined twin, and dies still dreaming. Which is to say, not all teams are guarded the same, which should be obvious enough.
Anyway, Chart 2 plots team ball movement to their percentage of open looks. As a general rule, team's that move the ball quicker generate more open shots. While the mission might be to try to stick on Korver's ass, over 80 percent of Atlanta's shots are open, in spite of being one of the league's best perimeter shooting teams. Their ball movement (about 2.5 seconds per touch) is part of the reason why.
At the other end of the chart, we see Minnesota, Brooklyn, Toronto and the Lakers holding the ball much longer per touch and generating some of the lowest percentages of open looks. The Lakers take the place of greatest dishonor, with the second slowest ball movement and the lowest percentage of open looks (at 55 percent).
Not every team holds tight to the trend in Chart 2. The Knicks represent the greatest underperformance. They have the third-quickest ball movement, an average lower than Atlanta, but only 66 percent of their looks are open. New York's problem is geometric. Phil Jackson is trying to squeeze a triangle, which has been inflated through years of successful implementation in an analytics-averse NBA and with elite talent, into Dolan's square, which has been deflated through years of suspect (and occasionally scandalous) management and mediocre play. Conspicuous inefficacy on the part of the Knicks is expected by now, and at least explicable.
On the other hand, no outlier breaks the trend like the New Orleans Pelicans. They play with the league's slowest ball movement, but generate the second highest percentage of open looks. It's bizarre, but true. Let's see how they do it.
First, the Pelicans' high percentage of open looks applies to both their 2- and 3-point attempts. See Chart 3. The Pelicans are about 8 percentage points higher in regards to both shot types.
Chart 4 demonstrates that the Pelicans' phenomenon is not the result of any one particular player's percent of open looks. Ryan Anderson, Luke Babbitt, Anthony Davis, and Eric Gordon are all seeing a higher percentage of open looks than the league average.
To better understand how the Pelicans can generate such a high percentage of open looks with so little ball movement, we're going to have to dip into their offensive strategy.
This season, NBA.com added a new feature to their stats page. In addition to videos of particular plays, visitors can now view "movement" accounts. These animations are generated from SportVU's spatial tracking coordinates. For an example, go to Anthony Davis's stat page here, and click on a blue stat link (such as the one showing his FGA per game). You will be prompted with choices. Click on "movement."
The movement animations show a top-down view of the court, with players marked by colored dots. The yellow dot is the ball. We'll pause here to explain how a normal sequence looks and the decisions that go into making it so, but feel free to skip ahead to the next section if you're just interested in what the hell's going on with Anthony Davis and the Pels.
In Image 5, the red team is Golden State and the blue team is Atlanta. When this animation was frozen, there were 9.4 seconds left on the shot clock and Paul Millsap (#4) had the ball.
Notice that in Image 5, the paint is wide-open. The Warriors are not showing much help defense from the weak side. This defensive formation is inviting Paul Millsap to drive.
There are a number of factors that must be considered when trying to determine the optimal position for a weak side defender. With the above example, the Golden State defenders must be weighing Millsap's driving ability. According to NBA.com's player tracking stats, Millsap is shooting 45.1 percent on drives this season. They must also consider the defensive ability of the player guarding Millsap. In this case, that is Steph Curry (#30). Curry had just switched with Draymond Green (#23) on a screen. Draymond Green would have been the preferred defender of Millsap for Golden State.
But a possession is a fluid thing, with defenders making on-the-fly decisions about not just whether or not to help or switch, but their resting defensive position on the floor—do you cheat a half-step closer to the lane because of Millsap's ability to put it on the floor, or do you stay wide and guard the three-point line—and all of this is informed by the specific strengths of your opponents. On this one, Millsap, who shoots 46 percent on drives, drove the lane and when defenders not covering Kyle Korver slid down to help, Millsap kicked the ball out to Teague for an open perimeter shot (he missed). This is the anatomy of a typical open shot; a player forces the defense to make a decision, and hopefully does it fast and well enough to create an opening for a good shot. But some players create bigger openings than others.
We now return to the Pelicans. If we're going to talk about ball movement and touch time, we should touch on another thing we can track using that data: touch time before a shot. Below is a chart that plots the average time of all offensive touches against the percent of shots that result in touches that took fewer than two seconds (presumably, most of these are catch-and-shoots and a result of a good possession). The basic idea here is that we want to see which teams tend to get shots off of the bounce and which get shots off of the catch.
We would suspect percentage of FGA on less than 2 seconds of touch time to correlate with percentage of open FGA. So, we would expect New Orleans' anomalous behavior in Chart 2 to appear again in Chart 6. But that's not the case! Instead, the Pelicans' touch time numbers are directly in line with the Lakers, their minimal ball movement brethren. While not what we might have suspected, the touch time statistics start to shed some light on New Orleans' open shot numbers.
The low percentage of shot attempts after less than 2 seconds of touch time suggests that the Pelicans feature at least one ball-dominant perimeter player that likes to create his own shot off the dribble. In fact, the Pelicans' Tyreke Evans leads the league with 768 total drives this season, and 601 of Tyreke's 944 FGA (64 percent) this season have come from less than 5ft.
Tyreke's drives account for some of the long-touch shots and also help draw help defense away from perimeter shooters. Tyreke's play partially accounts for the Pelicans' minimal ball movement and high number of open looks, but for the rest, we turn our attention back to where it truly belongs, Anthony Davis. It turns out that over 72 percent of Davis' long 2-point attempts are open. (The league average is 59 percent.) While New Orleans' offense probably creates some of these opportunities for Davis, his ability to attack the hoop is probably the biggest contributing factor. It's hard to imagine many power forwards or centers closing out hard on a mid-range Davis jumper. Doing so would almost certainly result in Davis blowing by the defender for a thunderous dunk.
When Davis is taken out of the equation, 64 percent of the remaining Pelicans' 2-point attempts (from > 10ft) are open. That number is still above league average, but far closer to reasonable given the teams' low ball movement.
We have partially solved the Pelicans' mystery. However, we do not yet know how the Pelicans have managed such a high percentage of open 3s. Davis has only taken nine 3-pointers this season. His only make was highly contested (and amazing).
Ryan Anderson is responsible for 324 of the team's 1248 3PA. Over 90 percent of Anderson's 3-point attempts have been open. Anderson is a true stretch 4. As such, he often draws larger defenders that may not be so quick as to recover on kick outs to Anderson, which leaves the big man open more often.
If Anderson were the only Pelican to see an above average percentage of open looks from 3, we might conclude that Anderson's rare combination of size and 3-point shooting ability is the answer. However, Anderson is not the only Pelicans player to see significantly above average open looks from the perimeter. The rest of the team is still seeing over 87 percent open looks from 3 (when the league average is 80 percent).
Over 90 percent of shooting guard Eric Gordon's 3PA have been open, even though he is shooting 48 percent on these attempts. Over 85 percent of Jrue Holiday's 127 3PA have been open. And even though Luke Babbit is shooting 50 percent from behind the arc, and 3-point shooting is almost his entire offensive repertoire, 90 percent of his 3PA have been open looks.
With so little ball movement, how are all of these Pelicans getting so many open looks from deep? This is where you see the true effect of Anthony Davis.
Few NBA players are as challenging to guard as Anthony Davis. The Pelicans occasionally run offensive sets where they get Davis the ball at or around the foul line. (According to NBA.com, Davis gets about 7 touches around the foul line per game.) This offensive design puts one of the most dynamic players in the game in a position where he has a plethora of options. If the defender gives Davis space, he can knock down the 15-foot jumper. If the defender plays tight, Davis can drive left or right. If the defense helps, Davis has the ball one quick pass from any perimeter shot, including the corner 3s. The situation is nothing short of a nightmare for defenses.
Image 7 shows how some defenses attempt to stop Davis's possessions at the foul line. Davis has the ball at the foul line in this game against Philadelphia. It's as if the rim is a special treat to be saved for a later feast, and Davis is Vacu-Sealing the defenders. Above, we talked about the many factors influencing the optimal position for help defense. Apparently, Anthony Davis with the ball in the center of the half court trumps all.
Below is another example of this working, this time against the Cavs. After setting a screen, Davis gets the ball at the foul line with his defender behind him. The Pelicans have 2 shooters ready and waiting on the 3-point line. Center Omer Asik, who is not a threat from the perimeter, hides on the baseline near the hoop, where he is still close enough to finish a layup if his man goes to Davis, but far enough away to give Davis some space to attack. Good luck Cleveland.
The effectiveness of these plays might lead you to wonder why the Pelicans are scrapping it out for the last playoff spot in the West, at least beyond the cosmic imbalance of conferences. Well, it turns out that while Anthony Davis has super powers, and those can affect a lot of good on an offense, just as often the New Orleans offense itself uses those powers for, uhm, it might be easier to just show you.
In the play below, Davis has the ball at the foul line with over 14 seconds on the shot clock. That's a great start. Luke Babbitt (who is shooting 50.9 percent on 3s for the season) is waiting in the corner. It would be foolish for Babbitt's defender to leave Luke wide open for a corner 3. So, Davis will have one less help defender to worry about should he decide to drive. The play looks fantastic so far. Half the court is set up perfectly. The other half…not so much. 3 Pelicans are grouped on the edge of the paint. This brings 3 defenders to a position where they can both guard their man and help on an AD drive. So what was the plan? Two of those 3 in the huddle set a screen for the third (Jrue Holiday). Then, Davis dribbled over for a dribble-handoff to Jrue. All of this was to get Holiday a 22-foot 2-point jumper.
Let's go through that again. New Orleans got Anthony Davis, one of the great players in the NBA, the ball in an incredibly dangerous spot with plenty of time on the shot clock. Then, they exploited this fortuitous position by running a play for a Jrue Holliday 22-foot 2-pointer. Holliday shoots about 36 percent from that area on the court. That's a play where the first option was a shot that generates .72 points per possession. In this case, Holiday made the shot, but still.
So that's about the size of it. The Pelicans are a unique team built around a unique player, and as often as that turns out for the best, they just as often put his unique skills to uniquely stupid use. Anthony Davis might well be unstoppable. But someone's really got to stop the Pelicans.
Image via AP
Stephen Shea is a professor of mathematics at St. Anselm College in Manchester, NH. He earned a Ph.D. in mathematics from Wesleyan University, and a B.A. in mathematics from The College of the Holy Cross. His writing has been featured in the Journal of Quantitative Analysis in Sports, in Psych Journal, at WinthropIntelligence.com, at TeamRankings.com, and at NBA Australia. Stephen is the author of Basketball Analytics: Spatial Tracking (2014), and a coauthor of Basketball Analytics: Objective and Efficient Strategies for Understanding How Teams Win (2013). Find more of his work at his blog, BasketballAnalyticsBook.com.