Author: Chibuzo Ikonte (Weinberg ’20)

Over the last couple of seasons, people have started to regard James Harden as a top five basketball player. It is easy to concur with that assertion when one looks at his remarkable regular season averages of 29 points, 8 rebounds, 11 assists, as well as his 22 triple doubles on the season. However, despite James Harden’s astounding nightly box scores, I have still remained skeptical while watching his game that he deserves to be considered an elite player in the NBA. As a result, I decided to meticulously watch and dissect his play this past postseason to try and find evidence to support my contention. Furthermore, while watching Houston’s series against Oklahoma City, it came to my attention that although Harden was putting up phenomenal averages of 33 points, 6 rebounds, and 7 assists, I found his numbers, especially his scoring output, to be inflated. This was in large part due to how often he drew fouls, especially on three point shots. In fact, during the regular season (as of March 21) Harden lead the NBA in drawing fouls on three point shooting attempts (via fivethirtyeight.com, BIGDATABALL).

Furthermore, during the series I also noticed that Harden was often one of the last people to get back on defense; it almost seemed as if he did not care enough to hustle back. (via fivethirteight.com and BIGDATABALL)

Also, based off of average speed on defense (via NBA.com) I also found statistical evidence to support my claim that Harden is the slowest defender in the league.

On top of this, during the series, I also noticed that James Harden turned the ball over a lot and took a lot of bad contested jumpers. Overall, I just felt that considering all of those things, I could not find myself calling Harden an elite NBA player, let alone a top five NBA player. However, my thoughts based on the eye test of Harden’s game are beside the point of this article. For example, someone could point to how much responsibility he has on offense holding onto the ball(usage rating of 35.3%) or Houston’s uptempo offense (6th in the league with pace of 99.05) to explain his turnovers. It follows then that critiquing Harden’s strengths and weaknesses are not the appropriate measure to answering my question of interest. In my opinion, rather than relying upon subjective observations and finding data to confirm my bias towards Harden, I felt that I needed to gather quantifiable advanced statistics that are better indicators of what constitutes a great player, and compare Harden’s output of that with other NBA “elite players.”

For my sample size of players to juxtapose with Harden, I defined an elite NBA player as someone named to the 1st team all NBA roster in the last six seasons. Also, it is important to note that the data that I will be gathering will be strictly based on postseason output, because an elite NBA player should be judged on what they do in the postseason, more so than the regular season. (Note: Anthony Davis in 2017, and Kobe in 2013 will be excluded from the data gathered since they did not participate in the playoffs the season that they were also named to the all-NBA first team). Player impact estimate, player efficiency rating, usage rate, defensive rating, offensive rating, plus minus, postseason win shares, value over replacement, and true shooting percentage are the variables that I considered to be more telling of an elite NBA player; this is in contrast to the rudimentary statistics often found in a box score.

Here is the output below:

Player impact estimate (PIE) is a statistic that essentially measures a player’s overall statistical contribution against the total statistics in games they play in. It uses the formula (pts+fgm+ftm-fga-fta+dreb+(.5*oreb)+ast+stl+(.5*blk)-pf-to)/(Gmpts+Gmfgm+Gmftm-GmFga-GmFta+GmDreb+(.5*GmOreb)+GmAst+GmStl+(.5*GmBlk)-GmPf-GMTO). Defensive rating measures a team’s points allowed per 100 possessions. “On a player level this statistic is team points allowed per 100 possessions while he is on the court. This statistic is also more exact than a typical player “possession based” calculations, because it calculates rather than estimate player’s’ possessions on the court. The same goes for the offensive rating statistic. Win-shares measures the estimated number of wins contributed by a player. True shooting percentage measures a player’s shooting efficiency taking into account twos, threes, and free-throws. Value over replacement is a box score estimate of the points per 100 team possessions that a player contributed above a replacement level player, translated to an average team and prorated 82 game season.

Here, I ran a t-test, testing the hypothesis that Harden’s PIE<18.37, the average for the “elite players” in the last six seasons. Thus, I ended up finding statistical evidence at the 5% level, that Harden’s PIE of 16.9 is less than the average elite NBA player. Furthermore, I ran the same test for player efficiency, plus minus, win-shares, true shooting percentage, value over replacement, and offensive rating and got the same results. At the 5% significance level, there is statistical evidence to say that Harden, when juxtaposed with other elite NBA players, is a below average player in those categories.

For Harden’s usage rate, I ran a t-test, testing the hypothesis that his usage rate was greater than 29.78, the average for an elite NBA player. However, after running my test, I found that there was not even statistical evidence, at the 5% level, to say that Harden is used more than our other subjects. However, in Harden’s defense, after running a hypothesis test for his defensive rating, I found that despite the liabilities that I pointed out in his defensive game, there is not statistical evidence at the 5% level that his bad defense affects his team more than other elite NBA players.

Thus, while there are many great things that James Harden does, such as generate a league high of 27 points from assists per night, I have found statistical evidence to show people that we need to slow down in our glorification of Harden. People should not continuously be fooled by his great scoring output, because after conducting my research, I have found statistical evidence to support my theory that James Harden is overrated; at least when juxtaposed with other elite NBA players.

Email: ChibuzoIkonte2020@u.northwestern.edu