NBA Free Agent Value Found Lurking on Defense: Data Viz

Lev Akabas
·5 min read

After a statement Miami Heat win in Game 5 of the 2020 NBA Finals cut the Los Angeles Lakers’ series lead to 3-2, Lakers coach Frank Vogel made a surprising decision to move point guard Alex Caruso into the starting lineup for the first time all season for Game 6. It paid off.

Caruso is not an offensive force; he averaged just 5.5 points per game on middling shooting percentages last season. But he is a defensive specialist, and a good one. In 2019-20, he ranked fifth among qualifying NBA players in steals per 36 minutes and finished in the top 10 at his position in every well-known defensive advanced metric.

It was a shtick last season for the Lakers’ bench players to appear shocked every time Caruso, an NBA player with a 36-inch vertical jump, makes an athletic play. What they actually should be shocked about was that, in the 2019 offseason, the team signed Caruso for just $5.5 million over two years.

Players like Caruso, who move the needle on defense despite relatively minor contributions on offense, generally earn much less than other players of a similar overall caliber. By analyzing free agent signings over the past three NBA offseasons, the data reveal that front offices undervalue defense relative to offense when it comes to cutting paychecks.

To assess overall offensive impact, we averaged each player’s offensive component of four reputable advanced metrics that estimate impact on team points scored per 100 possessions: RPM, PIPM, BPM and RAPTOR. We then did the same for the defensive side of the ball using the defensive components of those statistics.

This methodology isn’t flawless, but an average of four metrics with varying formulas paints a fairly complete picture. For example, using these averages, the three best defensive players in the NBA last year were Rudy Gobert (two-time Defensive Player of the Year), Giannis Antetokounmpo (2020 Defensive Player of the Year) and Brook Lopez (second in the league in blocks). The three best offensive players were James Harden, Damian Lillard and Luka Dončić, who all finished in the league’s top ten in both scoring and assists. Even if the metrics fail to accurately capture the value of any individual player, they’re sufficient for measuring overarching trends.

The chart below shows every free agent contract of the past three years plotted by annual salary and the player’s average offensive impact in the season prior to free agency, assuming he played at least 30 games. As you can see, better offensive players generally get paid more.

Though you’d expect a similar trend to exist for defense, it doesn’t. Some of the league’s best defenders are paid like role players and some of the highest paid players aren’t impactful at all on the defensive side of the ball.

The results of a statistical test back up the visuals. Our linear regression model suggests that an increase of a full point on our defensive impact scale (the difference between last year’s 165th and 75th best defenders in the league) yields just a 10% increase in salary, whereas a one-point increase on our offensive impact scale yields a 43% increase in salary.

Another way to look at this is to divide players into “offensive-leaning” (i.e. players with better offensive impact than defensive impact relative to the league average for their position) and “defensive-leaning” (vice versa). Let’s start by analyzing only “quality starters” (i.e. top 100 in the league in overall impact, but below All-Star level). In each of the past three offseasons, quality starters received significantly more money in free agency if they were offensive-leaning players than if they were defensive-leaning.

The same holds true for players lower down the ladder (i.e. between 101st and 240th in the league in overall impact). Once again, even within the same tier of talent, being a better offensive player rather than a defensive player will net you a better contract.

The Minnesota Timberwolves’ D’Angelo Russell may be the best example of a player at the other end of the spectrum from Caruso. He’s a gifted offensive weapon who can score in bunches and led the Brooklyn Nets to a playoff appearance in 2019 by buoying their subpar team offense. His slow lateral foot speed and light frame, however, have prevented him from providing that same value on the other end. In his past three seasons he’s finished 483rd, 499th and 512th in the NBA in defensive RPM. That last number is out of 512.

The 24-year-old Russell is a recent All-Star and former No. 2 overall draft pick. The 26-year-old Caruso is an undrafted bench player. In no universe would Caruso sign a more lucrative contract than Russell. Yet, despite the fact that Russell received a max contract in the same offseason that Caruso signed his deal at under $3 million per year, impact metrics indicate that Caruso was the better all-around player in 2020.

The chart above shows every free agent from last offseason. Players at the top are better offensively, players to the right are better defensively and players farther away from the line towards the upper right are better overall. Larger circles represent larger salaries.

The group of players in the shaded region—defensive-oriented, but impactful non-stars—is chock full of guys who played key minutes on playoff teams last year, including Caruso, Lopez, Wesley Matthews, Daniel Theis, Patrick Beverley and Dwight Howard. Many of them were signed for bargains compared to their more offensive-oriented counterparts.

What gives? Perhaps subpar teams with money to spare are more likely to spend it on offensive-minded players who can generate highlight reels and put fans in seats. Maybe front offices believe that defense can be easily improved through conditioning and increased effort, whereas it’s harder to teach natural feel for the game on offense.

Whatever the reasons, with free agency looming this weekend, we can identify some players who may represent a market inefficiency this offseason, and others who may be in line to be overpaid for their offensive abilities.

Alex Caruso ultimately played the fifth-most postseason minutes on the Lakers during their title run. This weekend, maybe next year’s NBA champion will steal an important contributor for a discount price simply because that player happens to be stronger on the defensive end of the floor.

More from