Part I: Which NBA and WNBA players are statistically identical?

This is a three-part essay that uses statistics and proprietary data to explore how the NBA and WNBA’s biggest stars align, why (despite their similarities) they’re paid differently, and why their differences in style and strategy is something to celebrate. Ultimately, my hope is that readers will come away with a better understanding of the inequalities and gendered differences that underpin professional basketball.

The author is me, Josh Strupp — product designer at Taoti Creative by day, data hobbyist the rest of the time.

  • Outlier Status: A glance at superstardom.
  • Superstar Similarities: Which NBA and WNBA players are (nearly) statistically identical?
  • Systemic Biases: Economies of scale, participation gaps, and more red flags as seen through chess, the MLS, and eSports.
  • Stylistic Disparities: how the WNBA’s strategy is better for fans and a data-driven look at “hero ball.”
  • In Conclusion: Thoughts on the latest WNBA collective bargaining agreement, a plea to engage with the sport, and a Kobe quote.
  • Glossary: includes listing of all abbrevitions used for stat categories (e.g. USG%, PER, PTS, etc).
  • Spreadsheet: Includes all 30,000 player comparisons and economic data leveraged to write Part II.

Outlier Status: A Glance at Superstardom

Note: This section is a crash course on outliers, focusing on how we can use stats to determine which players are most dominant. If you’re a basketball fan, and you semi-obsessively check your favorite players’ stats like I do, feel free to skip this section and jump to the next section: “Superstar Similarities: Which NBA and WNBA players are (nearly) statistically identical?”

They’re the mega-athletes that athlete better than the other athletes. They do front flips over entire teams. They’re LeBron James, Serena Williams, Lionel Messi, and Tom Brady. You know the type.

They don’t receive superstar status based purely on theatrics and wins, though that helps. YouTube highlight (or lowlight) reels tell half the story. As any sports fan would tell you, they also have to put up the numbers. Statistically, they need to be outliers — they need to have more points, make more defensive plays, and prove with undeniable, cold hard metrics that they are in fact bona fide 🌟s.

Within the realm of basketball, although the WNBA has about 10% of the NBA’s viewership, they have some standouts, many of which dominate in ways eerily similar to their NBA counterparts.

Anthony Davis is a freakin’ superstar. He was a first overall pick in the NBA draft in 2012. He’s a seven-time All-Star. And as of 2020, he’s got a ring.

Natasha Howard is also a freakin’ superstar. She was 2019’s Defensive Player of the Year. She was second in the league in blocks. She too played (and dominated) in the 2019 WNBA all-star game. And as of 2020, she’s got a second ring.

The numbers back this up. In 2019, Natasha Howard dropped 21.7.¹ points per game (PTS); that’s the most for the Seattle Storm by 7 points. Anthony Davis had an average of 25.9 during the NBA’s 2018–2019 season, which makes him the highest scorer on the New Orleans Pelicans (Davis was traded to the Los Angeles Lakers in 2020) by a margin of 4 points. Howard had 9.8 total rebounds (TRB) to Davis’ 12, which, again, puts both in the top spot on their teams.

¹ All WNBA numbers throughout this essay were adjusted based on variations in game duration — WNBA games are 8 minutes shorter than NBA games. Note that all statistics referenced represent the WNBA’s 2019 season and the NBA’s 2018–2019 season.

Below is a percentage breakdown of how much one player accounts for any given stat on Natasha Howard’s team (the Seattle Storm) and Anthony Davis’ team (the New Orleans Pelicans). For example, if one player takes two thirds of a team’s three-pointers, they’ll have a percentage of 66% in the 3P category. You’ll notice that Natasha Howard and Antony Davis are number one for most critical stat categories like field goals (FG), two pointers (2P), free throws (FT), and blocks (BLK). These are the cold hard metrics needed to prove you’re dominant.

Let’s zoom out to the rest of the league — like Anthony Davis and Natasha Howard, is there a small cohort of players that out-perform their teammates, or go above and beyond to full outlier status?

Below is a scatter plot containing every active² NBA and WNBA player from 2018–2019. Of the 393 (114 WNBA, 279 NBA) players plotted, fewer than 30 players shot more than 15 field goals (combined two and three-pointers) and made more than 8 of those field goals per game. That’s 7% of the entire population accounting for nearly 20% of all field goal shots and attempts.

Put another way — if there were 100 postal workers, seven “all-star” collectors would account for three houses each, while the remaining 93 mailmen and women accounted for less than one home (.86) each.

Not surprisingly, superstars like Brittney Griner, James Harden, and Giannis Antetokounmpo are hanging out in that top right quadrant, highlighting their offensive dominance. Hover over any data point to see where players land.

Note that there are more blue dots than orange dots due to the number of players in the NBA, which is roughly 4x that of the WNBA. This data set uses all players that played at least 2/3 of the season, which includes 393 total players (114 WNBA, 279 NBA).

Field goal metrics are generally a good measure of offensive superiority. To see how players defend, let’s look at the same population but measure steals (STL) and defensive rebounds (DRB).

Again, as you’ll see below, the data points become more sparse as you enter outlier territory. Interestingly, NBA players tend to have more rebounding outliers — Joel Embiid, Andre Drummond, Russell Westbrook, and Deandre Jordan are far outpacing the rest of the NBA and WNBA on the boards. The WNBA has outliers too, but their defensive play style and skillset lends itself more to stealing — players like the aforementioned Natasha Howard, her teammate Jordin Canada, Nneka Ogwumike, and Alyssa Thomas create more turnovers than any other players in either league.

I swear there’s a point to all of this preamble. This project began with a simple question — of the dominant players, which ones play the most alike? Style and showmanship aside, using pure numbers, which players contribute to their teams in the same ways? Is there a Brittney Griner of the NBA? Is there a Lebron James of the WNBA? Which players dominate similarly?

My starting with Natasha Howard and Anthony Davis wasn’t a coincidence. They aren’t just distinguished athletes, but when compared to every other player, they’re extremely alike.

Interesting, especially considering Davis makes 230 times what Howard makes ($27,000,000 vs. $117,000 as of 2019). More on the wage gap later.

² Active means the individual played at least two thirds of the season. For the WNBA this is 114 players; for the NBA it’s 279.

Superstar Similarities: Which Players are (Nearly) Statistically Identical?

In order to find statistical correlations between the WNBA players and their NBA counterparts, I built a stat-matcher, the inner workings of which I’ll explain below.

According to the matcher, Natasha Howard and Anthony Davis are an 86.69% match, which is high, especially considering that they play different positions.

For reference, here are Anthony Davis’ “stat match scores” when compared to Natasha Howard and two of her teammates: Mercedes Russell, their starting center, and Courtney Paris, their backup center. The higher the bar (i.e. the closer to 100), the more correlated the stats. Hover over a bar or a player’s name for readability.

Note: you’ll encounter a lot of abbreviations for stats (e.g. 2P%, PER, USG%. etc) throughout the essay. I will explain many of them, but feel free to use this glossary.

Natasha Howard is a forward, while Davis, Russell and Paris are centers. And yet, the overall stat match score between Davis and Russell is 11 points lower than the Davis-Howard score; Davis-Paris is 30 points lower.

The stat matcher shows that Russell and Paris fall well short of their superstar teammate in most categories, but they correlate with Davis in certain areas — for example, the rate at which a player is fouled and ends up shooting free throws (Ftr), or the percentage of possessions that end in a steal by the player (STL%). These correlations are likely due to Davis, Russell and Paris’ shared position. For those that don’t know, centers are closer to the rim, which means they’re prime targets for fouls and generally take a back seat to their smaller teammates who tend to steal the ball more.

The notable fact here is that despite their different positions, pretty much every per-game stat (FG through PTS going left to right) is highly correlated between Howard and Davis, plus critical stats like USG% (percentage of team plays utilizing a player while he or she was on the floor) and DWS (the number of wins contributed by a player due to their defense). This strong statistical resemblance reveals their matching star power.

But where do these numbers come from? How are they calculated? Do they take into account nuanced differences between the leagues? Why should I believe in these calculations?

Below is an explanation of how these scores were generated. The stat-matcher compares every active NBA player to every WNBA player, resulting in over 30,000 player comparisons generated using a Python script (see a sample for yourself). Apologies in advance for my long-windedness. Feel free to skip if you blindly trust me 💖

So, alphabetically by last name, the matcher:

  1. Takes Steven Adams (starting center for the Thunder in 2019, Jason Momoa look-alike) and compares him to Natalie Achonwa (Indiana Fever center, started about half the season, looks nothing like Jason Momoa) in 40 statistical categories. That is, to use one stat as an example, it looks at Adams’ two point attempts (10.1 per game) and compares it to Achonwa’s two point attempts (6.62 per game).
  2. The stat matcher first adjusts Achonwa’s number to account for the difference in game duration — NBA games are 8 minutes longer. So Achonwa’s two point attempts increases from6.62 to 7.68 per game. It’s worth noting that the NBA season is 48 games longer, so games (G) and games started (GS) were also adjusted.
  3. It then generates a similarity score, or stat match score, in each stat category. See the visual below, but in short, the score is created by taking the absolute value of Adams’ stat minus Achonwa’s, subtracting that number from the maximum difference (i.e. the most in one category minus the least), and dividing by the same max. This takes into account statistical outliers (looking at you, Brittany and Steph) and gives us a clean percentage. So Adams and Achonwa have a 86.54% correlation for two point attempts (roughly the same), a 25.77% correlation in free throw efficiency (Achonwa is significantly better), a 77.91% in total rebounds, etc. They have an overall stat match score of 51.50%, which is a weighted average of all stats.

4. The matcher then repeats the process for the next person on the list. So it compares Steven Adams to Rebecca Allen in 40 categories, then Kristine Anigwe in 40 categories, then Ariel Atkins, etc. Once Steven Adams is compared to all 114 WNBA players used in this data set, the matcher moves on to the next NBA player — the Miami Heat’s Bam Adebayo — who is then compared to Natalie Achonwa, then Rebecca Allen… you get the idea.

Note: You can download the spreadsheet and explore all 30,000+ comparisons yourself. Look for the sheet labeled “Stat Match Scores, Weighted Avg.”

For my NBA fans reading this: think about your favorite player. Think about his qualities and what he does better than anyone else. What you may not realize is he has a female doppelgänger (well, statistical doppelgänger) out there running the floor in nearly identical ways.

Take Kawhi Leonard. He’s a two-time NBA Finals MVP, and his team won the 2019 NBA championship. [Quick digression for non-fans: his wingspan is something to behold. So is his laugh.] Anyway, he is widely known as one of the most talented players in basketball offensively and defensively.

His WNBA stat match counterpart is likely to be dominant in a similar way — balanced, efficient, and quietly heroic. Sure enough, his most correlated superstar is DeWanna Bonner with an overall score of 89.02%. A veteran of the game entering her 11th season, she too is a (2x) WNBA champ, a 3-time all star, and has a colossal wingspan.

For reference, let’s see how Bonner’s stats compare to two other outstanding NBAers: Stephen Curry and Giannis Antetokounmpo. Steph, a guard notorious for his three-point shooting, and Giannis, a power forward known for his strength and presence at the rim.

You can see how, for the categories above, Leonard and Bonner frequently fall between Giannis and Steph or outperform both. This would indicate balance — dominance offensively, defensively, and beyond.

So who are Stephen Curry’s and Giannis Antetokounmpo’s WNBA matches? Use the dropdown below to see matches for Lebron, Elena, K.D., Arike, and more. What you find might surprise you. Hover over any bar to see an exact stat match score.

Note: For all you Sue Bird fans out there, remember that she was injured for much of the 2019 season and was therefore omitted from this data set.

Some interesting things to note below. As a reminder: WNBA stats have been adjusted for game duration (NBA games are 8 minutes longer).

  • While Steph Curry and Leilani Mitchell appear to have a low match score in three-point categories, Mitchell was actually tied for second in the WNBA for three- pointers made. Like I said, the stat matcher takes into account both league averages and outliers. This underscores Curry’s outlier status, as he scored significantly more threes than anyone in the NBA and WNBA.
  • 4x NBA champ Lebron James and 2x WNBA champ Kristi Toliver are alike in many ways. The volume of shots attempted (FGA, 2PA, 3PA) and shots made (FG, 2P, 3P, PTS) is not one of them. That’s because Lebron is taking twice as many (10 vs. 5) shots per game. Despite this clear display of “hero ball,” their offensive efficiency is very correlated (FG%, 2P%, 3P%, eFG%, TS%, OWS, and Wsper, or an estimate of the number of wins contributed by a player). As a reminder, find the stat abbreviation glossary here.
  • Arike Ogunbowale and Donovan Mitchell are extremely aligned, with a overall stat match score of 95.28%. The one place they fall below 85% correlation is DWS, or defensive win share. This is an estimate of the number of wins contributed by a player due to their defense. Arike’s team —The Dallas Wings—had a rough year in 2019. Defensively, she plays nearly identically to Donovan Mitchell, but the Utah Jazz actually won games.
  • Brittney Griner has led the league in blocks for the last seven years. It’s astonishing. Her stat match counterpart is LaMarcus Aldridge. She just about doubled his per game blocks, hence their dip in multiple blocking categories. Same thing with Jusuf Nurkić (1 BLK) and Jonquel Jones (2.4 BLK).
  • The aforementioned Natasha Howard stole the ball more than anyone in the NBA by a significant margin. This explains the massive categorical disparity with her closest counterpart Karl-Anthony Towns.

If Natasha Howard and Anthony Davis had a 86% match, then who got the closest to 100%? Why, Ian Clark and Kaleena Mosqueda-Lewis of course!

…I don’t know who these players are either. But fun fact: they were already mentioned in this essay. The very first chart revealing stat distribution for the Seattle Storm and New Orleans Pelicans included our beloved Howard and Davis, as well as their teammates Kaleena Mosqueda-Lewis and Ian Clark. Total coincidence. They are a 99.34% overall match.

Below are the top 500 correlations, sorted by the average across all stat categories. The last column includes their overall stat match score (scroll right for more stats). You can search by team, position, or player (note that if a player does not appear after searching in the table, it means they fall outside of the top 500, or under a 94% match. Numbers have been rounded for readability. You can explore all 31,000 comparisons here).

While doing research, I revealed some of my stat match findings to peers, mentors, and colleagues for feedback. Most said the same thing: look at the wage gap.

It’s especially notable (and unsettling) once you reveal how players can play the same game the same way, to remember that they receive astronomically dissimilar compensation. As reminder: as of 2019 Anthony Davis makes $27M a year; his stat match comparator Natasha Howard makes $117,000.

In America, on average, a woman makes 89% of what a man makes, despite having the same amount of experience and holding the same position. In the case of Davis and Howard, she’s making 0.43% of what he makes.

The average salary of an NBA player is $7.7M. In the WNBA, it’s closer to $75,000, so the female athletes are making about 1% of the salaries of their male counterparts.

These numbers are, of course, far from the whole story.

writer / creative director/ data scientist / corrupt politician /

writer / creative director/ data scientist / corrupt politician /