Dictionary.com tells us that the word “overrated” means: to rate or appraise too highly; overestimate
The word “high” is a relative term. To argue that something is “too high” or “too low” requires a reference point. Actually, to be precise, you need two points of reference.
For example, batting average in baseball will overestimate a player’s contribution if that player fails to draw walks and/or can’t hit for power. If we wish to argue that batting average overestimates a player’s contribution we must note where a player is ranked in terms of batting average and also where he is ranked in terms of a measure that considers plate discipline and power (i.e. OPS). In sum, we need two metrics to make the “overrated” argument.
Batting average has been calculated in baseball since the 19th century and it remains the most cited measure of a hitter’s effectiveness (one only needs to watch a baseball game on TV to see this point). When we turn to basketball, the “batting average statistic” (i.e. oldest and most cited) is points scored per game.
As we note in The Wages of Wins (and/or in this forum), this single metric does a remarkably good job of explaining the player evaluations of the coaches, general managers, and the sports media. And just like batting average, points scored per game is obviously not a perfect measure.
Focusing just on scoring ignores other factors that contribution to wins (such as rebounds, turnovers, steals, etc…). Scoring totals can also be inflated with more minutes and/or more shot attempts. In other words, lofty scoring totals may not coincide with scoring efficiency. Despite these shortcomings, scoring totals still dominate the discussion of a basketball player’s value.
Three Perspectives on the Overrated
The topic today is “overrated”, and as noted above, we need two reference points to tackle this subject. And given that this is The Wages of Wins Journal, the second reference point we will consider is Wins Produced. Specifically – as detailed in Table One — we are going to first look at how players rank in terms of both Wins Produced and points-per-game.
Table One reports the 15 players who are the most overrated by scoring. At the top of the list is Richard Jefferson. Of the 129 players who played at least 2,000 minutes in 2007-08, Jefferson ranks 9th in points per game. As detailed a few days ago, Jefferson – who used to be quite productive – doesn’t offer much beyond scoring these days. Consequently, with respect to Wins Produced he ranks quite low (106 out of 129).
Although points-per-game is frequently cited, there are of course more sophisticated metrics that try and take into account more of what a player does on the court. One of the oldest (and again, commonly cited) of these metrics is NBA Efficiency. This measure is calculated as follows:
NBA Efficiency = Points + Rebounds + Steals + Assists + Blocked Shots – All Missed Shots – Turnovers
Table Two looks at the most overrated players according to this measure.
Again our second point of reference is Wins Produced. And again we see Richard Jefferson topping the list. Joining Jefferson on both lists are Rudy Gay, Stephen Jackson, Rashard Lewis, Kevin Durant, LaMarcus Aldridge, Jamal Crawford, and Joe Johnson.
The similarities between the two evaluations are not surprising. The correlation coefficient between these two metrics for the 2007-08 season is 0.89 (NBA Efficiency per game and points scored per game has a 0.93 correlation).
Of course NBA Efficiency is not very sophisticated. This metric essentially adds together a player’s positive actions and subtracts off the negative. What happens if we turn to a measure that attempts to weight each player’s actions?
The most popular weighted measure is John Hollinger’s Player Efficiency Rating (PER). The simple version (okay, much simpler) of the PER metric – Game Score – indicates the basic weights employed.
Game Score = Points + 0.4*Made Field Goals – 0.7*Field Goal Attempts – 0.4*Free Throws Missed + 0.7*Offensive Rebounds + 0.3*Defensive Rebounds + Steals + 0.7*Assists + 0.7*Blocked Shots – 0.4* Personal Fouls – Turnovers
For 2007-08, Game Score and NBA Efficiency have a 0.99 correlation (this is basically what we find in other years as well). PER is a per-minute metric, and it has a 0.99 correlation with Game Score per-minute. Given the high correlation between these measures, we shouldn’t expect much difference in our list of overrated.
Before we get to the results, though, we need to note that since PER is a per-minute measure our second reference point has to be WP48 [Wins Produced per 48 minutes]. Despite this change, we still see some familiar names. Once again we see Richard Jefferson, Rudy Gay, LaMarcus Aldridge, Kevin Durant, Rashard Lewis, and Jamal Crawford. Jefferson, though, does not lead the list of overrated when we compare PER and WP48. Now LaMarcus Aldridge tops the list.
The Most Overrated
Although Aldridge tops the PER list, he is not the most overrated if we consider all three rankings. If we add together the difference reported in each table the most overrated player in the NBA for 2007-08 is….. yes, Richard Jefferson. Here are the 15 most overrated players if we consider all three perspectives.
1. Richard Jefferson
2. Kevin Durant
3. Rudy Gay
4. Stephen Jackson
5. Jamal Crawford
6. LaMarcus Aldridge
7. Michael Redd
8. David West
9. Rashard Lewis
10. Ron Artest
11. Allen Iverson
12. Tracy McGrady
13. Carmelo Anthony
14. Corey Maggette
15. Gerald Wallace
One should be clear that “overrated” does not necessarily mean “bad.” West, Iverson, McGrady, Anthony, Maggette, and Wallace were above average players [in terms of WP48] this past season. These players just weren’t as good as the scoring based metrics suggest.
In the next week (if I find the time) I will write a column on the most underrated. For those who want more on this topic, you can see a discussion of past seasons in Chapter 10 of The Wages of Wins. For more on NBA Efficiency, PER, and Game Score, please see the following posts:
NBA Efficiency: Do We Overvalue Rebounds? (November 9, 2006).
PER: A Comment on the Player Efficiency Rating (November 17, 2006)
Game Score: Marvin Williams Makes a Hypothetical Deal (December 16, 2007)
Our research on the NBA was summarized HERE.
Wins Produced, Win Score, and PAWSmin are also discussed in the following posts:
Finally, A Guide to Evaluating Models contains useful hints on how to interpret and evaluate statistical models.