August 7, 2023
Our Analytics and Insights squad has been working on a new feature to help basketball teams develop winning strategies – learn about how we are finding a player’s top comparisons.
Player comps provide a shortcut for coaches and scouts who want to learn more about a new player with whom they are not familiar. By comparing the unknown player to a more familiar one, they can quickly get a rough idea of how he plays. The most popular use for player comps is during the NBA Draft when collegiate prospects draw comparisons to the NBA stars they might try to emulate. For example, an NBA comp for the sharp-shooting lottery pick, Jordan Hawkins, could be Golden State Warrior legend, Klay Thompson (they share a comp score of 88).
To find someone like Hawkins’ top comps, we search for other players who have comparable stats across three broad categories: production, play style, and efficiency. We measure a player’s production with the basic box score using per-game stats for points, possessions, rebounds, blocks, steals, and more. We also consider style of play, the way a player tries to help their team score, based on various shot-type rates (like three-point attempt rate, ie. the percentage of their FGA that come from behind the line) and play-type rates (like their post-up rate, ie. what percentage of their possessions are used in the post). Finally, we look for players with similar efficiency based on their shooting numbers (3P%, FT%, etc.) and point-per-possession scoring averages (both on and off the ball). In the end, we’re using 33 different stats to find players with the most closely matched statistical profiles and then describing the accuracy of the match using a comp score from 1 to 100.
For example, here’s how the stats line up for Euroleague star, Vasilije Micić, and his top NBA comp, Spencer Dinwiddie (they share a comp score of 93).
Micić will be joining the Oklahoma City Thunder next season fresh off a Turkish Basketball Super League championship run with his club Anadolu Efes during which he garnered Finals MVP honors. He’s also a recent Euroleague MVP and back-to-back Euroleague champ. In the chart above, you can see the areas where both he and Dinwiddie were exceptional last season at their respective levels of play; both were in the 90th percentile or above (ie. they surpassed 90% of their peers) in on-ball scoring rate and playmaking rate (measures of play style), three-point shooting (a measure of efficiency), and possessions per game (a measure of production or “usage”, basically).
Top NBA comps for Micić also include two other dynamic ball handlers, D’Angelo Russell (they share a comp score of 87) and Zach LaVine (comp score of 87 as well).
Sasha Vezenkov is another Euroleague MVP who will be looking to make the jump to the NBA next season where he’ll be joining a talented Sacramento Kings roster looking to rekindle the fire of last year’s Beam Team. Like Micić, Vezenkov won his domestic league last season (The Greek A1 Basketball League) with his team, Olympiacos Piraeus, collecting MVP honors along the way. His NBA play-a-like is Lauri Markkanen, last year’s Most Improved Player. Both bigs can score efficiently while stretching defenses out to the 3-point line with more-than-capable shooting touch.
It’s worth pointing out, here, that these comps are *NOT* the same thing as career projections. We’re not predicting that Jordan Hawkins will become the next Klay Thompson or that Micić and Vezenkov will become NBA All-Stars like Russell, LaVine, and Markkanen. Because, first of all, our comp formula doesn’t account for age, height, wingspan or any measures of athleticism that might be linked to future success. And, moreover, the statistical similarity of two comps indicates only that they are outpacing (or falling behind) the peers in their own leagues by the same amount, not necessarily that they are directly comparable to each other. That is, while Vezenkov may have been the Markkanen of the Greek Basketball League last season, obviously that doesn’t mean we can expect him to achieve the same level of performance in the NBA next year.
Players who have gone back and forth between the NBA and EuroLeague, like Dante Exum, give us a unique opportunity to evaluate the comp-finding process.
Exum played six seasons in the NBA, mostly as a reserve, before joining the Spanish club FC Barcelona for a year and subsequently playing for Partizan Mozzart Bet Belgrade last season. Interestingly, during the 2020-21 season (his last in the NBA) Exum’s top comp was Alex Caruso and last year (while playing in the Adriatic League) his top NBA comp was Austin Reaves, the player who replaced Caruso on the Lakers.
To extrapolate a statistical profile from one league to another and project how an incoming international player might perform in the NBA, we need to know the “exchange rate”. We can get a sense for what NBA front offices think about the relative value of players overseas based on the market they demand for their services.
This summer, NBA free agents generally got deals that were in line with the salaries of their comps. I took a survey of 30 of the league’s latest hires – comparing the salary reported for each newly signed player to the median salary of his top 5 comps and found that free agents were getting about 110% of their comps’ comp (representing a very reasonable 10% raise given the continued growth of the salary cap). However, this was definitely not the case for newcomers Micic, Vezenkov, or Exum, who will all be making less than their NBA play-a-likes next season.
There are a lot of other reasons that a free agent and his new team might agree to a contract at a price point that is “not supported by the comps”. Cap space, for example, can be a huge factor in determining how big the checks are when they get cut. The Indiana Pacers ($75M, 26th in the league) and Houston Rockets ($44M, 30th) were two of the teams with the least amount of money on their books at the start of the summer and thus they had the most cap flexibility heading into free agency.
The result of all that cap space (and, indeed, the consequence of the constraints imposed by the minimum team salary of $122M) was that the Rockets and Pacers splashed around some cash this summer with big deals for players like Fred VanVleet (3 years, $129M), Dillon Brooks (4 years, $86M), and Bruce Brown (2 years, $45M). Each of these three contracts could be characterized as above expectations (at least relative to the salaries owed to each player’s comps for next season) but they made sense for these teams.
On the other end of the spectrum, the Phoenix Suns had one of the most bloated cap sheets in the league heading into free agency – with almost four times as much money ($164M) on the books for next season than the thrifty Rockets. Impressively, Phoenix cobbled together an entire bench worth of new players at the bargain price of nothing more than a collection of minimum contracts.
All of these new Suns have comps who will be making more money than them next year. In particular, the median salary for the top comps of Yuta Watanabe (2 years, $5M), Drew Eubanks (2 years, $5M), and Keita Bates-Diop (2 years, $5M) will each be in the $7 to $8 million per year range. Veteran players like Eric Gordon have a clear incentive to join a championship contender like the Suns and even younger players like Watanabe, Eubanks, and Bates-Diop would prefer to play for a winner.
In addition to the useful insight player comps offer NBA front offices and agents, we’re hoping our new tools will also be helpful to coaches who are preparing for- and game planning against their opponents. For college men’s and women’s teams we will be providing top Division and Conference Comps to simplified those efforts. Coaches and scouts can explore these and other types of player comps on the Synergy team site now, and we’ll have more to say about how, specifically, we’re calculating the comp scores in future posts, so check back here soon!
The development of the player comps project was led by Jonathan Lewyckyj. Jonathan is building tools to help coaches, scouts, and players find winning team strategies as part of Synergy’s Actionable Analytics Team.