The Best NBA Center in my Students’ Lifetime

Posted on September 7, 2008 by


My students this semester look to be about twenty years old.   This means that they were born in the late-1980s, or sometime around the time when I was a college student (and yes — as I rapidly approach forty — that makes you feel a bit old). 

After writing my last post – comparing Hakeem Olajuwon and Patrick Ewing (the latest Hall-of-Fame centers) – I started to wonder: Who is the best center in the lives of my students?

If start the clock sometimes in the eighties we have a few candidates.  But I think two stand above all others. 

The first center I am thinking of has the following characteristics:

  • More than $250 million in career earnings
  • Nine times named to the first team All-NBA team
  • One MVP award
  • Four times played for the NBA champion

These same characteristics were as follows for the second center:

  • About $118 million in career earnings
  • Four times named to the first team All-NBA team
  • One MVP award
  • Twice played for the NBA champion

When we look at salaries, awards, and championships, it’s pretty clear the first center is number one.  But when we look at the Wages of Wins metrics. a different story is told.

Admiral vs. Shaq

Table One: Comparing the Career Averages of Shaquille O’Neal and David Robinson

Table One compares the career averages – across the box score statistics and Win Score – for Shaquille O’Neal (center #1) and David Robinson (center #2).  When we look at the individual stats, we see that Shaq offered more in terms of shooting efficiency from the field, points scored, rebounds, and assists.  The Admiral had the advantage in free throw shooting, steals, turnovers, blocked shots, and personal fouls.  When we put the whole picture together – via Win Score – we see that Robinson comes out ahead.

How does this difference translate into wins? 

Table Two: Comparing the Career Performances of David Robinson and Shaquille O’Neal

Table Two compares each player in terms of Wins Produced and WP48 [Wins Produced per 48 minutes].  Through the 2007-08 season, Shaq had played 3,401 more minutes in his career.  Consequently, the Wins Produced story is quite similar (each produced a bit more than 250 wins in his career).

When we turn to WP48, though, we see that Robinson has posted the better mark. How much better?  Had Shaq’s WP48 been equivalent to the Admiral’s, he would have produced 31.4 additional wins in his career. 

And by making one change to his game, Shaq could have achieved this production level.  For his career Shaq has hit 52% of his free throws.  In contrast, Robinson connected on 74% of his shots from the line.  Has Shaq matched Robinson’s efficiency from the line, he would have produced 32.9 additional wins in his career and posted a 0.365 career WP48.  Robinson’s career mark was 0.363, so we see that Shaq’s inconsistency at the charity stripe could be considered the one factor that held him back.

Perceptions of Robinson

Although Shaq was less productive, it’s still the case that he has been paid more and received more awards.   Part of the difference in pay can be attributed to Shaq playing in more recent years and generally playing in a larger market.  The awards, though, are probably a different story.

Table Two not only reports the performance of Robinson and O’Neal, but also the average performance of their teammates (or everyone else on the team not named The Admiral or Shaq).  To give some perspective to these numbers, in 2007-08 the average WP48 of the teammates (or non-stars) on an NBA team was 0.076.  In Shaq’s first two years – and in Miami in the first part of 2007-08 – his teammates were below average.  In every other season, though, O’Neal was able to play with above average teammates.  In fact in nine seasons his teammates WP48 surpassed the 0.100 mark.

Above average teammates was also the norm for Robinson, but only after Tim Duncan arrived in 1997.  Prior to Duncan’s arrival, Robinson’s teammates posted an average WP48 of 0.076. In other words, unlike Shaq, for much of his career Robinson did not have an exceptional team around him. As a consequence, Robinson’s teams did not win as often as the team’s employing Shaq.  And one suspects – like we saw with Kevin Garnett before he arrived in Boston – the failings of Robinson’s teammates dimmed the perceptions of the Admiral’s performance.

As I have noted in the past, the purpose of player statistics is to separate a player from his teammates.  In other words, the analysis of player statistics should prevent us from confusing the performance of the team from the performance of the player. And when we look at all the stats – including Shaq’s woeful performance at the line – it appears that despite the edge in championships, awards, and money, Shaq is not quite as productive as The Admiral.

Okay, the Admiral tops Shaq.   But across the last twenty years, is Robinson even the best Spur?  To answer this question, let’s look back at Table One, where the career statistical averages of Duncan are also reported.  As one can see, Robinson offered more in his career.  If we turn to Wins Produced we see that Duncan has produced 211.9 wins in 30,610 career minutes.  This works out to a WP48 of 0.332.  This is about what Shaq has done across his entire career, although we have to remember that Shaq’s career numbers are deflated by his last two seasons.  If Duncan insists on playing until the bitter end, his career numbers will also take a hit.

What about The Dream?

What about the subjects of the last post, Olajuwon and Ewing?  Ewing – who was still very good – offered less than Olajuwon, Shaq, Duncan, and Robinson.  When we look at Olajuwon, though, we see that he produced 272.1 career wins.  This mark bests the other four players we considered.  But Olajuwon played nearly 10,000 more minutes in his career than Robinson.  If we turn to WP48, we see that Robinson’s career mark of 0.363 easily tops Olajuwon’s mark of 0.295.

Of course, some might notice that Robinson didn’t keep playing until his productivity descended into the average range.  In other words, had Robinson – like Shaq and Olajuwon – played until he couldn’t play anymore, perhaps the Admiral’s career numbers would be lower.  Although this might be true, we should also note that Robinson not only was better across his career, he was better at the peak of his career as well.  To see this point consider how many times each of these players surpassed a WP48 of 0.400 or 0.0300:

  • Robinson: +0.400 in three season, +0.300 in 11 seasons
  • O’Neal: +0.400 in one season, +0.300 in 12 seasons
  • Olajuwon: +0.400 in one season, +0.300 in 9 seasons
  • Duncan: never surpassed 0.400, +0.300 in 8 seasons
  • Ewing:  never surpassed 0.300 or 0.400

This list reveals that Robinson, at his peak, surpassed the performance of any of the other centers we considered.  And that includes Olajuwon.

Let me close by considering an argument people often offer in evaluating basketball players.  Often when considering whether player A is better than player B, people look at how the two players performed against each other.  Although this approach might work in a sport like boxing or tennis, it’s not appropriate in basketball.  Basketball is a five-on-five sport.  What matters is how a player contributes to his team’s success, not how he performs relative to one person on another team.   And when we consider each player’s contribution to team success, it looks like The Admiral was the most productive center in the NBA since the 1980s.

– DJ

The WoW Journal Comments Policy

Our research on the NBA was summarized HERE.

The Technical Notes at provides substantially more information on the published research behind Wins Produced and Win Score

Wins Produced, Win Score, and PAWSmin are also discussed in the following posts:

Simple Models of Player Performance

Wins Produced vs. Win Score

What Wins Produced Says and What It Does Not Say

Introducing PAWSmin — and a Defense of Box Score Statistics

Finally, A Guide to Evaluating Models contains useful hints on how to interpret and evaluate statistical models.