The Wages of Wins Journal

Shaq is Still the Best

December 6, 2007 · 28 Comments

A milestone in Shaquille O’Neal’s career is rapidly approaching.  On Friday night the Heat will take on the Golden State Warriors.  And assuming Shaq plays (and that he plays Thursday night against Portland), this will be O’Neal’s 1,000 regular season NBA game.

Given this milestone, I thought it would be a good idea to briefly look back on Shaq’s career and ponder where his game is at today.  And not to give away my punch line, but my pondering is going to reveal that Shaq is still the best.  Before I get to that, though, let’s go back in time.

Fifteen years ago I had just finished my first year of graduate school.  At that point I had not spent any time measuring player performance in the NBA.  And the only people who thought I was “stupid” were my girlfriend (now my wife) and most of my graduate school professors.  Of course, now everyone has caught on to the secret they knew a long time ago.  But that’s a story for another day.

Reviewing Shaq’s Career

Today we are thinking about Shaq.  Fifteen years ago O’Neal was the first player taken in the draft.  And he immediately rewarded the Orlando Magic, his new employer.  In 1992-93 a young Shaq produced 21.7 wins with a Wins Produced per 48 minutes [WP48] of 0.339 (average WP48 is 0.100).

Across the next twelve seasons Shaq continued to produce.  His lowest WP48 – in 1995-96 – was 0.284.  In every other season — except 1999-2000 when his mark hit 0.428 – Shaq’s per 48 minute mark was between 0.300 and 0.400. 

Entering the 2005-06 season Shaq had a career WP48 of 0.344 with 236.2 Wins Produced. Yes, his career WP48 was quite close to his rookie performance. In sum, across the first 882 games of his career, Shaq was consistently an extremely good player.

And then in 2005-06 his performance slipped.  He was still good, just not as good. Specifically, his WP48 fell to 0.225 and for the first time in his career he failed to reach double figures in Wins Produced.  The story at the time was that O’Neal was hurt.  After all, he missed 23 games due to injury. 

In fact, when the Heat went on to win the NBA Championship – the fourth title Shaq won in his career – it appeared that O’Neal was still as good as ever.

Last year, though, his performance slipped again.  His WP48 was 0.128 and in 40 games he only produced 3.0 wins.  Again, though, with 40 games missed due to injury, it looked like Shaq’s health was the excuse.

Still the Best

And perhaps this excuse was valid.  This year Shaq is supposed to be healthy.  And after 17 games, as Table One reveals, he’s once again the most productive player on his team.  Yes, Shaq is once again the best.

Table One: The Miami Heat after 17 games in 2007-08

Okay, this is not quite the achievement it would appear to be. After Shaq, the highest WP48 on the Heat is offered by Dorrell Wright.  And Wright’s mark stands at 0.128.

Although Shaq is the most productive player on this team, his WP48 is only 0.150.  Given this level, it’s not surprising the Heat are a very bad team.  With Seattle winning on Wednesday night, only the Timberwolves have won fewer games. 

The Heat’s efficiency differential, and Wins Produced (which is derived from efficiency differential), tells us this team is not quite as bad as their record indicates.  A 4-13 mark indicates this team is on pace to 19 games.  But with an efficiency differential of -5.7, a final mark of 26 wins is predicted. 

For Miami fans, I am not sure 19 wins is much different from 26 victories.  Both marks are quite bad and suggest the 2006 NBA Champions are either at rock bottom or rapidly approaching that mark.

When we look at Table One, we see one major problem.  Dwyane Wade – the player who has led the Heat in Wins Produced each of the past two seasons – has only been about average this year.  If Wade returns to what we saw last year this team’s projection would rise from 26 to 37 victories.

Two Shaq Questions 

And that might indeed happen.  What may not happen is a return to the Shaq we saw in the past.  To see this, let’s consider two questions.

Where has Shaq slipped? and

Why has Shaq slipped?

Table Two answers the first question.   This table reports what Shaq has done recently and for his career, relative to the average NBA center.

Table Two: Shaq’s Recent and Career Performance

As Table One notes, Shaq’s shooting efficiency from the field remains high.  He’s also still blocking shots and continues to have problem from the free throw line.

But with respect to rebounds, turnovers, and assists, Shaq is not what he used to be.  His work on the boards has especially declined, with Shaq now grabbing 2.5 fewer rebounds per 48 minutes (relative to his career average before the 2005-06 season).

So we can see where Shaq has faltered.  Now for the big question: Why has Shaq slipped?

We could keep using the injury excuse.  But at some point we have to acknowledge something that, if not true today, must be true at some point.  Shaq must eventually lose out to Father Time.

Remember, Shaq entered the league in 1992.  Fifteen years later, few members of his draft class are still in the Association.  Alonzo Mourning – Shaq’s teammate with Miami – is logging 16 minutes per game with the Heat. And Robert Horry has been on the court for only twelve minutes this year (not per game, for the entire season) for the Spurs. Other than that, every other player (I think) who arrived with Shaq on draft night in 1992 is now out of the NBA.

But while most of his draft class is spending their time watching basketball, Shaq is still out there toiling for 28 minutes per night.  Certainly he is still, as he has been most of his career, the most productive player on his team.  But his productivity, in absolute terms, is not what it used to be.

Every once in awhile, of course, the Shaq of old might re-appear.  For example, back on November 23rd he scored 26 points and pulled down 14 boards against Houston.  His Win Score for that game was 18.5.  Unfortunately for the Heat, that was the only game this season that he posted a Win Score the eclipsed his career per game average of 13.4. 

Motivating Shaq

All of this may or may not be a matter of motivation.  Remember back in mid-November, Dwyane Wade called out the Big Aristotle.  

“Probably this year more so than any year I have been more vocal with Shaq, talking to him and trying to motivate him,” Wade told reporters on Thursday. “But the main thing is Shaq has got to be self-motivated. He has got to be willing and ready to do it.”

Whether Shaq is self-motivated, or he needs Pat Riley and/or Flash to provide motivation, may not be the point anymore.  There is going to be a time, and we may indeed have reached it, where Shaq will no longer be able to dominate a game. And all the motivation in the world is not going to change that fact.

Let me close by asking on last question: When should Shaq give up?  Many people argue that great players should try and leave when they are still great. I completely disagree.  We would all pay money to take the court in the NBA.  As long as Shaq has fun playing, he should do what Rickey Henderson did in baseball. Keep playing until they send you home.  Remember, this is just a game.  If Shaq wants to play, and they let him, he should play.  Let silly sportswriters worry about his legacy.

- DJ

Our research on the NBA was summarized HERE.

The Technical Notes at wagesofwins.com 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.

Categories: Basketball Stories

28 responses so far ↓

  • Patrick // December 6, 2007 at 2:30 am

    Excellent post again.

    The other night I watched the Lakers-Wolves game on FSN, and a funny quiz appeared. The audience was asked who the best laker of all time was. You could text message in your choice from:

    1) West
    2) Mikan
    3) Magic Johnson
    4) Kareem
    5) Kobe Bryant

    Yes, you’re reading correctly. Shaq wasn’t on the list. I had to laugh. Even if you aren’t a WP48 fan, that omission is just a tiny bit over the top in its ridiculousness.

    I was tempted to SMS in a message saying “Um…HELLO…SHAQ?!”, but didn’t, of course…

  • mrparker // December 6, 2007 at 8:02 am

    Magic has got to be the best laker of all time. Unless any of those guys is also 1b or 1a greatest player of all time.

  • Joey G // December 6, 2007 at 8:08 am

    A friend of mine was raving about Iverson’s play last night and I have to admit… I enjoyed watching Iverson light it up for three quarters too.

    But I pointed out to him that Allen may be good on your playground but he is only average at best in the NBA. After all the Nuggets lost.

    He misses too many shots! If he would have made the nine shots he missed, then the Nuggets would have won by 14 points instead of losing by 4. I told my friend to read your book and that missed shots are bad and rebounds are good.

    I said that Marcus Camby was the star of the game because he had twenty rebounds. 20! This is better to do because you get lots of credit for getting the rebounds in the Wages of Wins and that’s how you win. By getting rebounds.
    And not missing shots like Iverson always does.

    But then my friend said that Camby scored 0 points and a team of Cambys would get 100 rebounds and end up with 0 points and lose 106 – 0.

    I said it doesn’t matter. If he just gets lots of rebounds and does nothing else, he’s a superstar even if his team of Cambys never scores a point.

    Isn’t that right?

    I mean, the Bulls were great because of Dennis Rodman weren’t they?

    A three-on-three tournament of three Rodmans would kick three Jordans butts because the Rodmans would get lots of rebounds and rebounds are good.

    Wait a minute… do you mean the Rodmans actually have to do something with the ball once they get it?

  • dberri // December 6, 2007 at 8:20 am

    Joey G (or from your e-mail address TG Randini),

    If Iverson always played as well as he did last night, he truly would be one of the top 50 players of all time.

    Is there a reason you switched names?

  • Owen // December 6, 2007 at 8:30 am

    Joey – Iverson had a pretty good game last night. He actually posted a higher PawsMin than Camby. But Kobe was better. Winscore of 13 in 30 minutes versus Iverson’s 15 in 48. Najera, Fisher and Vlad were also better on a PawsMin basis than Iverson…

    The Bulls were great because of Dennis Rodman. They were significantly better with Rodman than they were with Horace Grant. And when he was in Chicago, the Bulls were in fact the greatest team of all time. They wouldn’t have been without him.

  • Panda Bear // December 6, 2007 at 9:44 am

    “If Wade returns to what we saw last year this team’s projection would rise from 26 to 37 victories.”

    Wouldn’t it really rise to 35 b/c at 4-9 they have underperformed that rate by a couple games so far? Your 37 figure seems to project that future performance across the whole season, including the wins/losses already logged.

  • Oren // December 6, 2007 at 9:48 am

    I was noting that Camby’s twenty rebounds are considered better then Iverson’s 18-27 FG shooting(66%) and 15-18FT shooting(83%).

    On the one hand, Iverson shot a pretty impressive 66% on 27 shots. That’s amazing.

    On the other hand, if Iverson had missed say four more shots, then he would have shot only 14 for 27. That’s still good. But I don’t know if I’d say that it’s amazing. Even if he missed three of those shots that he hit, I don’t know if I’d say 15-27 is amazing.

    It shows me that there is less of a difference then I would have thought between what I consider a great shooting game and what I consider an ok shooting game.

  • Panda Bear // December 6, 2007 at 9:51 am

    Owne writes: “It shows me that there is less of a difference then I would have thought between what I consider a great shooting game and what I consider an ok shooting game.”

    It’s the same thing in baseball. The difference between a .250 hitter and a .300 hitter is just one hit per week. This is why statistics are good. It’s difficult to detect that difference just watching.

  • dberri // December 6, 2007 at 9:52 am

    Panda,
    My projection took what he did so far as given. So they get 37 wins if Wade turns into what he was last year from the next game to the end of the year.

  • Panda Bear // December 6, 2007 at 10:41 am

    dberri, Thanks for the clarification!

  • Animal // December 6, 2007 at 1:08 pm

    Very off topic question. Dberri, what do you think about head coaches? What kind of effect do you think they have on the game, if any. Do they in any way effect players and if so how does this affect your judgment of players

  • Panda Bear // December 6, 2007 at 1:18 pm

    Dberri says Don Nelson has the effect of making his players better: http://dberri.wordpress.com/2006/12/27/taking-requests-%e2%80%93-the-golden-state-warriors

    Also, JC Bradbury did work saying Leo Mazzone made pitchers’ ERAs drop by 0.5 runs. But that work was completely discredited by Mazzone’s disastrous oversight of the Baltimore pitching staff after he left Atlanta.

  • dberri // December 6, 2007 at 1:53 pm

    Panda and Animal,
    My co-authors and I have found that some coaches have an impact. Nelson was one of these coaches. Some coaches, though, do not have an impact.

    And Bradbury did find the same thing. Not sure I would agree that Mazzone’s experience in Baltimore discredits Bradbury’s finding.

  • Panda Bear // December 6, 2007 at 2:31 pm

    Given Bradbury’s model with the 0.5 ERA reduction in the original paper as well as what the variability of pitching staff ERAs are (using the latter to construct a confidence interval), for Mazzone’s staff to do so poorly in Baltimore should happen about as frequently as I should spontaneously turn into a real panda bear. I’d say discredited was a fair assessment!

  • Panda Bear // December 6, 2007 at 2:33 pm

    I do like the methodology and your result of Nelson being a good coach. I think your result there will be predictive out of sample. I’m not sure why the model failed in baseball. Maybe it’s b/c the Braves’ GM was just really good at finding pitchers about to improve and dumping them before they were about to deteriorate.

  • John // December 6, 2007 at 3:04 pm

    Re: Mazzone

    I think you’re thinking of it incorrectly, Panda Bear. If Mazzone has dropped his staff 0.5 ERA, that doesn’t mean the ERA is good, just that it’s half a run better than it would have been.

    For example, let’s say I have a staff of pitchers who should have a 6 ERA (we’ll call them the Texas Rangers), Mazzone would turn them into 5.5 ERA pitchers. But really, they’d still stink.

    Or, another way to think about it would be to say, Mazzone could reduce my ERA by half a run, but really, it still wouldn’t make me a major league pitcher.

    Not that I have any idea if this is actually happening or not, just thought I would point out that Mazzone coaching a bad team does not necessarily discredit the fact that he’s making them better. I think to actually discredit it you would have to do research on pitchers in Baltimore to figure out whether they pitched better before/after his time with them. Which, I believe, is the study that is being referred to above.

    Here’s the main link if you’re interested: http://baseballanalysts.com/archives/2005/03/the_mazzone_eff_1.php

  • John // December 6, 2007 at 3:07 pm

    Thought I should point out that the rest of that after the first paragraph is not for Panda Bear, who obviously _has_ read the study, but in case anyone else is interested (and partially because someone else already linked the similar entry Mr. Berri had on basketball coaches).

  • Kent // December 6, 2007 at 4:06 pm

    Panda Bear,

    Did you calculate if Baltimore’s ERA was relative to what it would have been with a neutral pitching coach? Or were you just assuming that because it was so bad Mazonne couldn’t have had a favorable impact?

  • Kent // December 6, 2007 at 4:08 pm

    Actually Mr. Bear, we can’t yet check because JC Bradbury’s calculation looks at performance both before and after the pitching coach’s time. As of now we only have the before time. It’s not looking good for JC’s prediction but if these pitchers continue to worsen than it’s not necessarily because Mazonne was bad influence.

  • Top Posts « WordPress.com // December 6, 2007 at 4:59 pm

    [...] Shaq is Still the Best A milestone in Shaquille O’Neal’s career is rapidly approaching.  On Friday night the Heat will take on […] [...]

  • Gerry // December 6, 2007 at 5:57 pm

    Shaq is slow… He’s done…

  • TG Randini // December 6, 2007 at 7:51 pm

    dberri,

    I boo-booed on a cut on paste for the name without looking. Didn’t mean to be a ‘Joey G’.

    I find it interesting that Iverson had 18-27 and 15-18 and someone noted he only had a ‘pretty good’ game. And Camby can sit under the basket and collect rebounds and be of similar value.

    Re Rodman: I loved the guy. Highly entertaining… and an excellent team player, defender and rebounder. He got 3 rings with the Bulls.

    But Horace did, too.

    What was the constant?

    MJ.

    (Or in his dream of dreams… Scottie.)

    I am sure you overrate rebounds (for the many reasons I’ve stated previously using some econ terminology) and underrate scoring. But so be it… it’s your system.

    I think you would be closer to the mark if you adjusted your win score formula which has ‘points – field goal attempts…’ to have a multiplier in front of the ‘points’ to account for the fact that avg. FG% is less than 50%. Then take what the multiplier adds to the points and subtract this from your rebounds via another multiplier so that the total correlation of the formula is the same.

    Your formula as presently constituted seems to imply that tossing up a FG attempt is like flipping a coin. But the problem is… the coin is imbalanced because the average shooter has less than a 50% chance at success.

    Always,
    TG

  • mrparker // December 6, 2007 at 8:52 pm

    tg,

    Doesn’t position adjusted win score account for the average shooting percentage of each individual position. So, in that respect players are only being graded on the average field goal percentage of their peers.

  • dberri // December 6, 2007 at 8:54 pm

    TG,
    Mr Parker is right. A player with an average level of shooting efficiency would be average in WP48 and PAWSmin. So what you propose is already happening.

  • dustin // December 6, 2007 at 11:17 pm

    Are jump balls kept track of in the boxscore at all? Do they count as turnovers if you started with it and ultimately don’t end with it?

  • Pete23 // December 6, 2007 at 11:19 pm

    Yahoo Sports has a great article on Jamario Moon:

    http://sports.yahoo.com/nba/news;_ylt=AhjG3ZQMPf.kL9rH5EHCSUm8vLYF?slug=aw-moon120607&prov=yhoo&type=lgns

  • Pete23 // December 6, 2007 at 11:20 pm

    Dusty, Jump balls are not tracked in the box score. HEre’s an example of a box score– http://sports.yahoo.com/nba/boxscore?gid=2007120606 Allen Iverson is on fire!

  • dustin // December 6, 2007 at 11:58 pm

    http://www.nba.com/features/conferencechallenge0708_bigeast.html

    ever think of doing something like that dave?

Leave a Comment