by Daniel Alpert
In 6th grade, I remember watching the NBA draft on TV. The Celtics had the last pick of the draft and took Semih Erden of Turkey. I remember being bummed. I know it was the last pick of the draft, but who the hell was Semih Erden? His highlight tape was lackluster and I had never seen him play.
New York Knicks fans share the same sentiment every time a European player is drafted. The receptions they gave Danilo Gallinari and the welcome to Kristaps Porzingis would be enough to scare anyone out of New York (although I’m sure they regret jeering Porzingis). Fan bases are always disappointed when European players are drafted. And while there are certainly cases of European players becoming MVPs (Dirk Nowitzki, drafted 9th in 1998) and being the steal of the draft (perennial all-star Tony Parker, drafted 28th in 2001), there are high profile cases of European busts (like Darko Milicic being picked at #2 over Dwyane Wade and Carmelo Anthony in 2003 and Andrea Bargnani being picked 1st overall in 2006).
I set out to find if European players drafted were actually worse than non-European players drafted or if it was just our bias that disappointed us every draft. Using data from the last 20 years (up to the 1996 NBA draft), I ran t-tests on the means of European and non-European players for several statistics. The statistics tested were: years played, games played (G), points per game (PPG), rebounds per game (RPG), assists per game (APG), minutes per game (MPG), free throw percentage (FT%), three point percentage (3P%), field goal percentage (FG%), win shares (WS), win shares per 48 minutes (WS/48), box plus/minus (BPM), and value over replacement player (VORP). Win shares is how many wins a player is responsible for in his career. BPM is a players approximate plus/minus (points for minus points against while on the court), and VORP is a metric determining a player’s value over a generic replacement. I compared the average European players to the average non-European taken in the same draft slot. For example, are European players taken at number 5 in the draft worse than non-Europeans taken at number 5?
The below chart shows the averages for all players from the last 20 years taken at each draft slot. A box highlighted red indicates that the European players had a statistically significant lower mean in that category than non-European players taken at that slot. A green highlight indicates that European players had a statistically significant higher mean than their non-European counterparts at the same slot. An unhighlighted box indicates there was insufficient evidence to accept that the two groups were any different. Lastly, for draft slots where there were not at least two European players draft, I highlighted that row gray and did not test on it, as the sample size would be too small.
Obviously the graph is smattered in red, showing that there are far more instances where European players were significantly worse than non-Europeans than vice-versa. In a vast majority of cases, the mean statistic for the Europeans was lower than the mean for the non-Europeans, but small sample sizes often led to the results not being statistically significant. While there were many instances where as many as 6 or 7 European players had been drafted at a given draft slot in the last 20 years, often times, there were only 2. These small sample sizes lead to means that could very well be anomalies and are therefore much harder to declare as significantly different than the means of the non-Europeans with larger sample sizes. Also, Europeans seem to perform better in advanced metrics than in normal counting statistics, which is consistent with the perception that they pass well and have good instincts, but are possibly less athletic or tough.
One thing that I did notice was that many players (especially in the later rounds) did not record much playing time at all. In fact, many of the European players drafted in the second round never played in the NBA at all, and that while these zeroes (0 PPG, 0 G, 0 FT%, 0 BPM, etc.) were affecting the means of both groups, they were clearly hurting the statistics of the Europeans more. I reran the same program, this time only including players who had recorded at least 100 career minutes (essentially players who had played about 5 games or more).
The results are below, with the same highlighters.
Adjusting for having actually played, Europeans stack up very well against non-Europeans. In fact, there is no discernible evidence as to which group fares better. Indeed, for Europeans that actually make the league, they are just as good as non-Europeans who make the league.
So why all of the European hate?
Most European players who are drafted are relatively unknown to their American fan bases. Certainly an aspect of this European hate is fear of the unknown. Who is this player? Is he really worth a draft pick? Can he make our team better? Another aspect is that fans want their team to draft players that they rooted for in college, watched play in March Madness, and have seen ESPN coverage about. Picking an unknown European player will annoy any fan whose favorite college player is still on the board.
While these first two reasons may not have merit, the last one does. As evidenced by the fact that I had to adjust the initial study, many drafted European players never end up playing. Of the 174 European players drafted in the last 20 years, only 98 have played at least 100 career minutes. Only 56% of drafted players touch the court for more than a fleeting second. Of the non-Europeans draft (1006), 819 or 81% get at least 100 career minutes. For this reason, European players often never come to fruition and therefore are a bigger risk to draft.
This still does not mean European players are worse. It could indicate that European players are often overdrafted. While still deserving to be picked by a team, maybe a European player should have been selected 10 slots lower than he actually was.
There are also many flaws in this study. Most noticeably, the sample sizes were very small for any given group of players. Also, in calculated the means, all players were weighted the same. For example, Kobe Bryant’s 25 PPG over his 20 year career was weighted the same as Kirk Haston’s 1.2 PPG in 134 career minutes in calculating the average PPG of the #16 overall selections.
When watching a game, you won’t notice any skill difference between European players and non-Europeans. On the court they are the same. The discrepancy lies in the mass of European players who never made it.