Spacing and shooting are cornerstones of modern NBA offenses. The art of the drive and kick has become essential. How can defenses stop the drive-and-kick?
Modern NBA offenses rely on spacing and 3-point shooting in a cascade of effects — placing shooters in corners and spacing others around the arc creates room for ball handlers to penetrate and get to the paint. Crafty ball handlers can drive into the paint to score or collapse the defense to pass out for an open 3. Many times the ball handlers are not looking to score once they get into the paint. Instead, they are looking to collapse the defense and kick the ball back out for a high-value shot.
This drive and kick mentality may be best-exemplified Bucks. Only in the modern NBA can Giannis get triple-teamed on a drive to the paint and kick it out for an open 3 to the 7-footer Brook Lopez. James Harden and the Rockets have perfected the drive and kick as well. So, how are the defenses supposed to counter the deadly efficiency of a good drive-and-kick?
One answer comes from a recent Wharton Moneyball episode, featuring an interview with Seth Partnow. As a former Director of Basketball Research for the Bucks, Partnow knows a thing or two about the advantages of the drive-and-kick. Seth mentioned that it might not make sense to help on the drive in many instances. An off-ball defender has two options when a ball handler drives to the basket; step in and help cut off the drive or stay home on his man. Most likely, if the off-ball defender goes to help, they will leave their man open for 3. If the open player is a competent 3-point shooter, the defense has given up a highly valuable shot to help stop a drive to the basket. The question is whether the drive to the basket is more or less valuable than an open 3.
To try to answer this question, I took data on drives and shots by zone from NBA.com. The goal is to compare the value of a drive and the value of an open 3. To make this comparison, I needed to calculate the expected points scored per drive for the driving ball handler and the expected points per shot (PPS) for his teammates. Calculating PPS for the teammates is straightforward; multiply the player’s field goal percentage for each zone by the value of the shot. I’m assuming most drive and kick opportunities are attempting to kick out for an open 3. Therefore, I only calculated PPS for the left corner, right corner, and above the break (ATB) 3-point attempts.
Calculating the expected points scored per drive requires a little more explanation. The assumption here is that the most valuable option for the offense in a drive and kick is the kick-out for an open 3. In an attempt to limit the advantages of the drive and kick, defenses must choose to take away specific actions of the offense. In this scenario, the goal of the defense is to take away the driving player’s option to pass out to an open teammate. Assuming the defense can take away the driving player’s option to pass, there are only three outcomes remain for a drive. The driving player can attempt a shot, draw a foul, or commit a turnover. The defense is trying to force a one-on-one contest against the ball handler and the on-ball defender.
Now that we have simplified the drive to only three options, we can calculate the expected points scored per drive. Expected points per shot, expected points from free throws, and expected points from turnovers are the possible outcomes of a drive. Technically, points from turnovers could have a negative value for the offense. But for simplicity, we will assume the value of a turnover is zero. That leaves only two components that make up the expected points scored per drive. The expected points per shot is the probability of making a shot on the drive multiplied by 2 — the value of a 2-point field goal. The expected points from free throws is the probability of making a free throw after being fouled on the drive. The formula is below.
Notice the probability for each component is just the specific outcome, field goals made (FGM) or free throws made (FTM), over the total possible outcomes of field goal attempts (FGA), fouls drawn (PF), and turnovers (TOV).
Now that we can calculate the points per drive and points per shot for we can look at some specific examples. First, let’s look at some of the most and least efficient scoring drivers. The figure below shows the top ten most efficient drivers followed by the least efficient scorers. The players are limited to playing in at least ten games and playing at least ten minutes per game.
We can compare these expected points scored per drive with each player’s teammates to determine which teammates are most dangerous to leave open on the drive.
Let’s start with the reigning MVP and efficient attacker of the rim, Giannis Antetokunmpo. Giannis, with expected points per drive of 1.12 would be an efficient offense all on his own. If the Bucks coaching staff wanted to be lazy, they could give the ball to Giannis every time and still be a top-three offense in terms of offensive rating. The graph below shows the 2019-20 Bucks roster compared with Giannis’ expected points per drive. Check out the interactive version here.
The graph shows anyone with a bar above the dashed yellow line is expected to score more points per shot than Giannis can be expected to score on his drive. If the bar is above the line, then the defender should stay on his man. If the bar is below the line, then the defender can leave his man in that area to help on Giannis. The graph shows the difficulty in defending the Bucks as the defense has to pick their poison. The big guarding Brook Lopez can’t leave to protect the rim, especially if Lopez is in the left corner, against Giannis without giving up a valuable 3-point look to Lopez. If Wesley Mathews or Kyle Korver is on the court, it makes it even harder on the defense as both of these new additions provide efficient scoring from deep. The best hope is to send whoever is guarding Eric Bledsoe or Pat Connaughton before Giannis can get into the paint. Which is how the Raptors held the Bucks offense in check in the Eastern Conference Finals this past year.
The Houston Rockets are another efficient offense based on spacing and shooting. With James Harden as their primary creator, the Rockets have one of the most efficient scorers of all time surrounded by deadly efficient shooters. Harden’s expected points per drive are slightly less than Giannis’ at 1.05 points per drive. However, the shooters surrounding Harden are more dangerous than the Bucks’ shooters. Defenses should never leave Eric Gordon or Danuel House if either of these players is on the court and behind the arc. Then there are P.J. Tucker and Gerald Green who can be expected to be more efficient almost everywhere behind the arc. A ton of shooters surrounding the most efficient one-on-one player in the league is a headache for opposing defenses. See Harden’s interactive plot here.
Finally, to look at a worst-case scenario, I chose to look at Lonzo Ball. Ball is the only player with starter-level minutes in the bottom ten of the expected points per drive. The graph below shows that defenses should never worry about helping on a Lonzo Ball drive. Every one of the players on the 2019-20 Pelicans can be expected to score more efficiently from behind the arc than Lonzo is expected to score on his drives. Here is Lonzo’s interactive graph.
Most players fall somewhere in between Giannis and Lonzo in terms of scoring efficiency on the drive. Most players have teammates in certain scenarios that are fine to leave and others where the defense should stay at home. It is almost always better to not help on the drive unless the driver is one of the best at getting to the rim or his teammates are horrible shooters. It seems in most cases, it is better to live with a drive and a shot than die by a kick out for an open 3-pointer.