Very early on Monday morning Henry Abbott – of TrueHoop – posted the following:

Congratulations, David Berri

June 15, 2009 1:29 AM

Champion of the 2009 TrueHoop Stat Geek Smackdown.

Berri, the lead author of the Wages of Wins, beat ESPN’s John Hollinger by seven points. Hollinger picked every series of the playoffs right, except for three — Houston vs. Portland in the first round, and both Conference Finals. Berri, meanwhile, correctly predicted that the Lakers would beat the Nuggets.

Soon afterwards my phone just started ringing.  Members of the media from around the world kept calling, wondering the secret to my victory.  And here is the story I told…

It all began when the Lakers acquired Pau Gasol.  At the time Andrew Bynum was still hurt.  But I argued that once Bynum was healthy, the Lakers with Kobe Bryant, Gasol, Bynum (as well as Trevor Ariza and Lamar Odom), would be the favorites to win the NBA title.  That was my argument in the midst of the 2008 NBA Finals (when Bynum was still hurt).  And that was what I argued at the onset of the 2008-09 season.

Now that the Lakers have indeed won the 2009 NBA Championship we can take a step back and evaluate my immense predictive powers.  Not only did I correctly identify the winner in thirteen of the fifteen 2009 playoff series, I also correctly identified the NBA champion in 2009 more than twelve months before it happened.  Such a result clearly indicates that what is said about basketball in The Wages of Wins is correct. 

Then again….

Let’s take a step back and identify two problems with the previous two paragraphs. 

1.  Okay, no one actually called me (not even Henry). 

2. Although I did pick the Lakers to win more than 12 months ago, one crucial detail was incorrect.  The key to the Lakers was going to be the return of Andrew Bynum.  Bynum the person did return to the court, but the Bynum we saw in 2007-08 did not.  In 2007-08, Bynum posted a WP48 [Wins Produced per 48 minutes] of 0.394 (average is 0.100).  This past season his WP48 was 0.198. Such a mark is quite good, but not nearly as good as what we saw in 2007-08.  Plus, Bynum was again hurt.  In 2007-08 he only played 35 games before a season ending injury.  This past season he only managed to appear in 50 games. 

When we look at Bynum we see a clear decline in performance.  This could be due to

• Bynum’s injury problems.
• diminishing returns (the team did add Gasol), although this effect tends not to be so large.
• the possibility that what Bynum did in 1,008 minutes in 2007-08 was not representative of what he will do over time in the NBA.

All three explanations might have some validity (although, once again, the diminishing returns story probably cannot explain the size of the decline we observe).  Regardless, I assumed Bynum would return to what we saw in 2007-08 and therefore concluded the Lakers were the clear favorite to win the title in 2009.  The Lakers did win, but I am not sure that would have happened if Orlando didn’t upset Cleveland.  In fact, I was quite happy to see that upset.  Picking LA to beat Orlando seemed much easier than trying to guess if Cleveland or LA would be victorious in the NBA’s Kobe-LeBron dream finals.

Learning from the Smackdown

So what have we learned from the entire TrueHoop Smackdown experience?

Picking playoff games is really not a very good test of a model or someone’s analytical skills.  As I have been saying all along (see HERE and HERE and HERE), a seven game series is too short for predictions to be made with perfect accuracy.  In other words, the best model can be done in by the randomness of a small sample.

That being said, I think we have some evidence for the elements of a “best” model (at least “best” under the circumstances of the playoffs).  Two years ago I lost to Justin Kubatko in this contest.  As I noted at the time, both of us made our picks according to a team’s efficiency differential.  Kubatko, though, considered home court advantage and I did not.  Because both efficiency differential and home court advantage matter (and this can be seen statistically), Kubatko had the better model and he probably deserved to win.

Last year I did not participate and Kubatko repeated as champion.  This year I was back and Kubatko was absent.  But with a model based on efficiency differential and home court advantage I was victorious.  So we have now seen three consecutive seasons where the person who only considered these two factors managed to win the contest.   Continue reading →

Here is an interesting factoid about the NBA Finals.  Since 1978 (the first year we can calculate Wins Produced) no team has won an NBA title without one regular player (minimum 41 games played, 24.0 minutes per game) posting at least a 0.200 WP48 [Wins Produced per 48 minutes].  Only one team – the 1978-79 Seattle Super Sonics [led by Gus Williams with a 0.208 WP48] – managed to win a title without a regular player crossing the 0.250 threshold. And only four other champions didn’t have at least one player surpass the 0.300 mark. This tells us – and hopefully this is not a surprise – that to be an elite team you must have at least one elite player.

Okay, now let’s connect this factoid to the draft.  Since 1995, no player who posted a below average college PAWS40 [Position Adjusted Win Score per 40 minutes] his last year in college managed to post a career WP48 above the 0.200 mark (after five seasons, minimum 5,000 minutes played).  So although college numbers are not a crystal ball (and really, college numbers are not perfect predictors of what a player will do in the NBA), it does seem like players who don’t play relatively well in college are not likely to become superstars in the NBA.

Now let’s apply these two pieces of information to the upcoming NBA draft. What do Jrue Holiday, Jonny Flynn, DeMar DeRozan, and Jordan Hill have in common?

1. These four players represent picks 7 through 10 in Chad Ford’s current mock draft.  

2. All four players posted below average PAWS40 numbers last season.

An average player drafted since 1995 posted a PAWS40 of 10.13.  Here is what this quartet offered last year:

Jrue Holiday: 9.17

Jonny Flynn: 8.64

DeMar DeRozan: 7.76

Jordan Hill: 9.95

 And when we look at picks 11-20 we see the following names and numbers:

Gerald Henderson: 9.70

Austin Daye: 9.23

Earl Clark: 8.53

B.J. Mullens: 7.74

Jeff Teague: 9.97

Sam Young: 8.33

These players were also below average with respect to PAWS40 last season.  And given what we have seen in the past, none of these players are likely to become superstars in the NBA.  So if Chad Ford’s latest mock draft is accurate, we have some evidence – before any of these players start playing in the NBA – that half of the first 20 players selected will not become NBA superstars.  And it is likely – before we ever see the broadcast on draft night – that at least some of these players will be touted as potential superstars when they are drafted.

One last note on the subject of superstars: Since 1977-78 there have been 848 teams.  Of these, only 216 – or about 25% — had a regular player with a WP48 beyond the 0.300 mark.  Another 183 teams – or another 22% — had a player with a 0.250 WP48.  So this means over half of all teams did not have one player that seems a prerequisite to win a title.  And it tells us that New York, Toronto, Utah, Phoenix, Chicago, Houston, Atlanta, Milwaukee, Philadelphia, Oklahoma City, Washington, Denver, New Jersey, Memphis, and Sacramento have at least one move to make if they wish to contend for the 2010 title.

- DJ

The WoW Journal Comments Policy

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.

Once again it is time for the NBA draft, that wonderful time of the year where we discover that there are suddenly an abundance of future stars just waiting to join your favorite NBA team.

Looking Back at 2008

To see this point, here is what Chad Ford – the draft expert at ESPN.com – had to say last year:

• #4 – Russell Westbrook: Overall, he has a chance to be a better version of Rajon Rondo.
• #16 – Marreese Speights: He is kind of a poor man’s Elton Brand.
• #23 – Kosta Koufus: He could be the second coming of Mehmet Okur — a sweet-shooting big man who can play inside and outside.
• #27 – Darrell Arthur: He has a chance to be an Antawn Jamison-type player.
• #28 – Donte Greene: Lots of scouts compare him to Rashard Lewis.
• #34 – Mario Chalmers: He’s kind of a poor man’s O.J. Mayo
• #41 – Nathan Jawaii: a huge player from Australia who looks like a bigger version of Elton Brand.

One year later here is what this collection of future “stars” has accomplished:

• Russell Westbrook: 4.0 Wins Produced, 0.071 WP48
• Marreese Speights: 1.6 Wins Produced, 0.063 WP48
• Kosta Koufus: 0.8 Wins Produced, 0.070 WP48
• Darrell Arthur: 0.2 Wins Produced, 0.008 WP48
• Donte Greene: -2.4 Wins Produced, -0.161 WP48
• Mario Chalmers: 6.6 Wins Produced, 0.120 WP48
• O.J. Mayo: 2.3 Wins Produced, 0.035 WP48
• Nathan Jawaii: yet to play in the NBA

An average NBA player has a 0.100 WP48 [Wins Produced per 48 minute]. Of these players, only Chalmers surpassed the mark of an average player.  To put that in perspective, here is how some of the players these future “stars” were compared to performed at the start of their respective careers.

• Rajon Rondo: 7.0 Wins Produced, 0.184 WP48
• Elton Brand: 9.7 Wins Produced, 0.155 WP48
• Mehmut Okur: 3.1 Wins Produced, 0.109 WP48
• Antawn Jamison: 4.9 Wins Produced, 0.220 WP48

Rashard Lewis was drafted out of  high school and only played 145 minutes his rookie season.  In his second season, though, he produced 4.7 wins with a 0.142 WP48.  

So we see the actual stars each produced from the start of their career.  The players identified by Ford, though, have so far struggled.

To be fair to these players, most rookies struggle.  And most rookies do not develop into NBA stars.  But this is not the story we hear when the NBA draft season is upon us.

Evaluating Jordan Hill

To illustrate this point, consider what Doug Gottlieb had to say about Jordan Hill (insider access required):

What I like: Has played basketball for only the past six years competitively. Hill has played hurt, played tough and loves a physical game. Hill, like Thabeet, is more used to getting his points off the rim and not off the pass, making it an easier transition as a team’s fourth or fifth offensive option.

What I don’t like: Is not great at any one thing, and seems more like Etan Thomas than Brian Grant.

Best case: A Brian Grant-type

Jordan Hill – a power forward out of Arizona – is generally thought to be a lottery pick (Ford’s latest mock has him going 10^th to Milwaukee). Although Grant and Thomas are not considered “stars”, each had productive seasons in the NBA. Across 12 seasons, Grant produced 60.3 wins and posted a 0.135 WP48.  Most of these wins were produced for Portland and Miami.  In seven seasons with these two teams, Grant produced 51.3 wins with a 0.167 WP48. 

Again, Gottlieb thinks Grant is the best case scenario for Hill.  And that doesn’t look to bad. But what if Hill is actually Thomas?  Thomas had had trouble staying healthy, but he has produced 17.3 wins with a 0.125 WP48 in his career.  His third season was the only time he managed to appear in more than 75 games, and that season he produced 5.6 wins with a 0.143 WP48.

So it appears the best case and worst case for Hill looks pretty good.  But did Gottlieb get this analysis correct?  One check is to compare what these players did in college.

The average power forward taken in the draft since 1995 posted a 12.5 Win Score per 40 minutes (WS40) his last year in college.  Last year, as a junior, Hill posted a 12.3 mark. So he was slightly below average.  Grant as a junior, though, posted a 16.4 WS40 while Thomas had a 13.7 mark.  In other words, both were above average as juniors (and each was also above average as seniors). 

Now Win Score in college is not a perfect predictor of NBA performance.  But players who are below average in the college tend not to develop into above average performers in the NBA.  So the team that drafts Hill is probably not getting an NBA star (or even Grant or Thomas).

Hill or Blair?

All that being said, it does appear that after Blake Griffin, Hill will be the next power forward taken.  This means some lottery team is going to pass on DeJuan Blair to take Hill.  Let me close this post with a quick comparison of these players.

Table One: Comparing Jordan Hill to DeJuan Blair

As Table One indicates, Blair offered more than Hill with respect to everything except free throw percentage and personal fouls.  Once again, performance in college is by no means a perfect predictor of what we see in the NBA.  And there are suggestions that Blair might have a problem with his knees.  But there is an immense difference between what these players did last year in college.  So even if Blair is a health-risk, there does appear a good chance that if he stays healthy will be a very productive NBA player (which is not the story the numbers suggest for Hill). Continue reading →

On Thursday I had the privilege of giving a seminar at BYU.  The subject was the next book and it was fun talking with the students about some of the stories we will tell. 

While I was having fun, though, the usual Thursday post in this forum was skipped.

To make up for this, here are a few items of interest (hopefully):

Superman vs. Shaq

Adrian Wojnarowski has written an interesting column detailing Shaquille O’Neal’s behavior towards Dwight Howard (and Kareem Abdul-Jabbar).  Apparently Shaq believes there is a substantial gap between Shaq and Superman.

In an effort measure the gap, here is a ranking of every player who has ever played for the Orlando Magic.

Table One: Ranking the Orlando Magic (1989-90 to 2008-09)

As one can see, Superman tops the list.  Fans of Shaq would note that Howard produced his wins in five seasons while Shaq only played four years in Orlando.  In Shaq’s fifth season, though, he only produced 13.3 wins and this would not be enough to close the gap.  Shaq did post a higher WP48 [Wins Produced per 48 minutes] in Orlando, but that’s primarily due to the fact Howard started playing at 19.  If we look at what each player did from the age of 20 to 23, the WP48 of each player is essentially the same (not that a 0.329 vs. 0.306 is really that different in the first place).  So it doesn’t look like Shaq can claim he was ever that much better than Howard.

Here are a few more observations from Table One.

• Tracy McGrady is ranked 4^th in the history of the Magic.  As one can see, once upon a time McGrady was a very productive NBA player.  That is no longer the case today.  With the Magic he posted a WP48 [Wins Produced per 48 minute] above 0.200 each season.  He has not done this for the Rockets since his first season in Houston.  And now that he is 30 years of age, we might suspect that McGrady is not likely to reach the 0.200 mark again.
• Nick Anderson is currently second on the list.  The next players on the list who are still active with the Magic are Hedo Turkoglu and Jameer Nelson.  Turkoglu would have to produced 58 more wins to catch Anderson, and given Turkoglu’s level of production and age, that seems unlikely.  Nelson could catch Anderson, but he is going to have maintain his current productivity and stay healthy. If that happens (and those are big ifs as Nelson ages), Nelson will catch Anderson in five seasons.
• Scott Skiles is currently ranked 10^th.  Nelson should pass him next season, but I am not sure Skiles is remembered for being an average point guard. 

Readers Explain Randomness

My last post on the random nature of the playoffs resulted in a number of comments that suggested the point of the post was being missed.  While I was getting ready to post a reply, though, readers jumped in with comments explaining the role randomness plays in the playoffs. Continue reading →

At the start of each football game a coin is tossed to determine who will receive the opening kick-off.  Let’s imagine if instead of just one team calling heads or tails the fans in attendance were also asked to make a call.  And let’s further imagine that if you make the correct call, you get to stay.  But if you are wrong, you have to leave.

Okay, now let’s do a bit more imagining.  Let’s say 80,000 fans are in attendance – and since fans know it is a fair coin (equally likely to be heads or tails) — about 40,000 make the wrong call.  So these fans exit the building.  After they are gone, let’s imagine we play the same game again.  This time, about 20,000 fans are incorrect and they depart.  And then we play it again, and again, and again…  After three tosses we are left with about 10,000 fans.  After seven tosses there are about 625 fans.  After twelve tosses we should still have about 40 people left in the stands.

Now what have these 40 people learned?  These people have just called a coin flip correctly twelve consecutive times.  Clearly these people are incredible at this game. 

If we play the game one more time, though, we should expect about 20 more to depart.  What will these departing fans have learned?  Well, clearly they just didn’t match-up with the 20 who got the 13^th call correctly.  And they better go home and figure out why that particular match-up didn’t work if they ever wish to see another football game.

The above scenario was adapted from Nassim Nicholas Taleb’s book Fooled by Randomness [(2005): pp. 165]. This book argues that people often have problems understanding randomness.  And what it says is relevant to how people see the NBA playoffs.

Orlando Better than Cleveland?

Consider the Eastern Conference Finals. Continue reading →