In the wake of a seven-game NBA Finals and the one of the most unpredictable drafts in recent memory, the NBA community turned off their cell phones this weekend, sat down on the couch with a bowl of chips, and finally got around to watching those Netflix DVD’s which had been sitting next to the TV since the All Star break. But after a weekend spent relaxing, recovering, and being disappointed by Hansel and Gretel: Witch Hunters, it’s time to get back to work. Free agency is here once again, and though this crop of free agents may not be as bountiful as years’ past, the stakes have never been higher for NBA teams.
The effects of the new CBA are finally being felt as teams scramble to sign impact players while trying to avoid the dreaded repeater tax. In addition to a more punitive luxury tax, the new rules have limited exceptions for tax-paying teams making it harder to sign valuable role players. Now more than ever, teams have to maximize player value for their contract dollar. One bad deal can cripple a team financially and fatally limit their roster flexibility for years to come (Unless of course Danny Ainge is willing to take those contracts off your hands.) Properly valuing free agents and having the discipline not to overpay for them is the name of the game this summer. In the spirit of that, I’ve spent the last several weeks building an economic model for player salaries. Welcome to the NBA Skills Market 2.0.
The principle behind the Skills Market is simple. Every player in the NBA has specific skills and abilities that make them valuable. Teams quantify that value with the annual salary paid to the player. Some skills, like scoring and shot blocking, are valued very highly, meaning teams are willing to pay more money for players who possess them. Other skills, like generating a lot steals, aren’t valued nearly as much. The Skills Market model achieves two things. First it determines which skills, stats, abilities, and career achievements have the most significant effect on player salaries. Second, it assigns dollar values to each of those criteria. These dollar values are essentially the current market price for a specific level of production or achievement. With these values in hand, current player salaries can be evaluated to determine whether a player is overpaid, underpaid, or properly paid. More importantly, these values can also be used to predict how much a free agent will sign for in the open market.
Before we get started, a few notes about the model.
I used salary figures from the 2012-13 season for every player except the handful of restricted free agents who had already signed an extension. For those players, I used their 2013-14 salaries. Players on rookie contracts, non-guaranteed contracts, and minimum contracts for the 2012-13 season were excluded from the model. Rookies have not yet had the opportunity to have their value determined on the open market. Because this is a model for free agents who will sign guaranteed contracts, I excluded the non-guaranteed guys. And minimum contracts were excluded because they can artificially inflate a player’s value simply because of the number of years he has spent in the league. For player stats, I went back four seasons to 2009-10. Stats were averaged on a per game not per season basis.
Criteria: > 20.0 points per game
Most of the stats I used are valued on a singular basis (Example: $ per rebound or $ per assist.) Points were the exception. As was the case in the original Skills Market last season, points scored correlated better with salary when tiered, not valued on a per point basis. It will come as no surprise that elite, 20 ppg scorers are valued so highly by teams. They are quite rare. Over the last four years of statistical data, there have only been 19 players who have averaged 20 or more points per game.
2nd Tier Scorer
Criteria: > 15.0 ppg
The similarity in value of Elite and 2nd Tier scorers is proof of how inflated the market for secondary players really is. There are far too many players in the NBA who are paid as though they’re elite, but who produce like 2nd bananas. The Brooklyn Nets will start five of them next season. It’s impossible to win a championship without a capable secondary scorer. But teams shouldn’t necessarily pay those players like they’re a perennial All-NBA selection.
3rd Tier Scorer
Criteria: > 12.0 ppg
Even if they don’t drop 20 points in a game all season, teams still need players who can score double digits and keep defenses honest. For many of the players in this category, their scoring production is just an added bonus because they do so many other important things for their teams like rebound and get easy baskets for their teammates.
Criteria: Isolation points scored per game (according to Synergy)
Value: $853,645 (per point per game)
One of the biggest improvements I wanted to make in this year’s model was to evaluate not only the amount of points a player scored, but also how he scored those points. I considered several different methods for answering that question. Eventually, it became clear that the most valuable scoring ability is being able to create shots for yourself. Synergy measures a players’ isolation scoring in points per play. I’ve extrapolated those figures out to assess how many points a player has scored in isolation per game over the last four years. As you can see, a player who averages just two points per game in isolation is nearly as valuable a player who averages 18 points a night (2nd Tier scorer.) More than 22% of Kobe Bryant’s value is derived from his 7.33 points per game in isolation (Kobe, no surprise, is the league leader in that category.)
Post Up Scoring
Criteria: Post up points scored per game (according to Synergy)
Value: $374, 293
The rationale for measuring the value of post up scoring is the same as isolation scoring. Having the ability to pass the ball to a player and have him score without needing a screen or any help from a teammate is a valuable skill. Every team needs at least one player who can catch the ball in the post and go to work. The per point value here is lower than isolation scoring, but that seems to fit with the decline of post play in today’s NBA. More than 43% of Nikola Pekovic’s current salary value (and 21% of his predicted free agent salary value) comes from his 5.41 post up points per game.
Free Throws Made
Criteria: Free throws made per game
Just like points scored in the post or isolation, points scored at the line are quite valuable. I looked at both free throws made and free throws attempted. FTM had a better correlation to salary almost entirely because of Dwight Howard who misses more free throws per game than 90% of the league attempts.
Criteria: Shots blocked per game
At first glance, the value per block seems astronomical. But consider the fact that the number of blocks a player like Dwight Howard has in the box score doesn’t capture the number of shots he changes or the way he causes offenses to alter their strategy of attack. For players who aren’t pure shot blockers, blocks are an indication of effort and positioning. The value for blocks is so high because much more so than steals, blocks are a great indicator of defensive impact and value. Players like Monta Ellis lead the league in steals but would never be mistaken for great defenders, where as guards who don’t patrol the paint, but still average nearly a block per night are generally good defenders.
Criteria: Rebounds per game
Criteria: Assists per game
These values don’t compared to points scored, but they are still the basis for many players’ value. Guys who can grab double digit rebounds or dish out a lot of assists can find themselves very well compensated for their production, even if they’re limited in other areas. A player like Reggie Evans derives nearly all his value from his work on the boards. Likewise, the Knicks’ point guard duo of Raymond Felton and new Brooklyn Nets’ head coach Jason Kidd, can attribute nearly all of their value to their assist figures.
All Star Selections
Criteria: All Star appearances in the last four years
I looked at a number of different honors including selections to the All Star team, All-NBA team, and All-Defensive team. All Star selections correlated the best to salary by a wide margin (I suspect this is due to sample size. Each season there are at least 25 All Star selections whereas the All-NBA teams select only 15 and the All-Defense teams select only 10.) Quantifying recent All-Star selections (last four years) was much more accurate than using total All Star selections. The idea being that recent All Star selections are more valuable than selections in the past.
Criteria: 3-point FG made per game
This valuation of 3-pointers looks incredibly low at first. When I initially ran the numbers, I was convinced I had made a mistake. But when I looked closer, it became clear just how easy it is to find good three point shooters. The market is just saturated with them. And the salaries that 3-point shooters garner are typically quite low. Of the Top 30 shooters in 3PM/game this season, nine were on rookie contracts. This speaks to teams’ ability to find cheap shooting through the draft. Also, 10 of the Top 30 shooters made $5 million or less (non-rookie contracts.) Players like Danny Green, Randy Foye, and C.J. Miles knock down two shots from downtown per game and create spacing for big men inside. But they do so on short term contracts with less than average annual salaries. The majority of these players’ value comes from their scoring overall. They’re ability to shoot threes is important, but it’s the most replaceable skill in the NBA.
This model is not perfect by any means, but it does help put into context what NBA GM’s value the most and what NBA player agents use to promote their clients. Come back tomorrow to see which free agents are in for a big time pay day in the NBA Skills Market.
All salary information was obtained from Sham Sports. The stats were culled from MySynergySports and ESPN.com. I’d like to thank Professors David Rapach and Muhammad Islam of the St. Louis University Economics Department for their help.
Names are slowly creeping in as teams fill out their summer league rosters. So far, it appears Iowa State shooting guard Chris Babb and Davidson big man Jake Cohen have been offered spots on the roster.