Five Questions with Dave Berri

Recently, I had the chance to exchange emails with noted basketball analyst Dave Berri.  Berri was one of few analysts who saw potential in the Bucks at the start of the season, pegging them to finish sixth in the East.  Berri’s new book, Stumbling On Wins, is available on now and will be in stores soon.

The Milwaukee Bucks in 2009-10
After 71 games (team has won 39 games)
WP48 = Wins Produced per 48 minutes
Data taken from
Return to The Wages of Wins Journal

[table id=10 /]

* – Player is either a rookie, or has limited (recent) NBA history, so 08-09 numbers are the same as 09-10
**- WP48 from 2008-09 is calculated relative to position played in 2009-10
*** – 2008-09 numbers listed for Delfino, Stackhouse, and Brezec come from 2007-08.  Ilyasova’s 2008-09 numbers come from 2006-07

Obviously the Bucks have been the surprise story all year, but you pegged them much higher than most before the season, what did you see differently then and what are you seeing now?

My answer begins with the data (which hopefully is not a surprise). The table presents two perspectives on the players employed by Milwaukee this year.  The first view is the productivity of Milwaukee’s players – measured via Wins Produced and WP48 [Wins Produced per 48 minutes] – this season.  We also see how the team would look if we assume that what the players did last year (or in their most recent NBA season) would be seen again this season.

The first view (columns 3-4) says the Bucks should have won 38.7 games this season (as of Friday night).  Had these players maintained what they did last year, though, the team would have expected – as of Friday night — to win 32.6 games (columns 5-6).  The only real difference is the play of Ersan Ilyasova.  Back in 2006-07, Ilyasova played very poorly in 973 minutes at the age of 19.  Now that he is 22 he has become a very productive player.

If we knew Ilyasova was going to play this well, then we would have suspected the Bucks – even with the injury to Michael Redd – would have been an above average team this year.  Even without knowing about Ilyasova, though, we still would have suspected the Bucks were better than the media projected.

Remember, the experts at expected the Bucks to rank 14th out of 15 Eastern Conference teams before the season started.  But Milwaukee had the following above average players on their roster (above average based on past performance): Andrew Bogut, Carlos Delfino (he has generally been a very good NBA player), Luc Mbah a Moute, and Luke Ridnour.  With four above average players it seemed unlikely – as I noted last September (see the post below) – that this team would be awful.

Revising Expectations Upwards in Milwaukee « The Wages of Wins Journal

To understand why the media’s projection was incorrect we need to understand that scoring tends to dominate the standard evaluation of NBA players.  The most productive players on the Bucks, though, are not scorers. Consequently, we should not be surprised that this team was under-estimated.

Aren’t Ridnour’s numbers something of an aberration this season?  Both his FG% and 3FG% are at career highs, and they’re significantly higher than they typically are.  He’s at an age where it seems, well, bizarre, that he’d make such a leap.

Ridnour is shooting better this season.  But his rebounds and steals are down.  So his overall production – although improved – hasn’t changed that dramatically.  You are correct that players don’t tend to improve at the age of 28 (the peak appears to be around 24 or 25).

Why has John Salmons been the perfect fit for Milwaukee after struggling this year with the Bulls?

Not really sure I agree that Salmons was struggling with Chicago.  Last season, Salmons posted a WP48 mark very close to average (average is 0.100).  With the Bulls his WP48 mark was 0.070, so he was not quite as productive as he has been with the Bucks (WP48 = 0.107).  But his production is close.  His adjusted field goal percentage is a bit better than what we saw in Chicago, but his rebounds, steals, and turnovers are down.  What does appear to has changed the most is his shot attempts.  This has led to more scoring and gives people the impression Salmons is making a larger impact than an analysis of his overall numbers would indicate.

One should note that Salmons is basically doing for the Bucks what he did in 2008-09.  Across his career, Salmons has been an average NBA player.  In other words, he hasn’t been bad.  But he has generally not been the most productive player on the teams he plays for, and that still seems to be true for Milwaukee.

Your thoughts on Ilyasova and Mbah a Moute’s success at the four since Warrick’s departure?

It does appear Mbah a Moute has played much better in the second half of the season.  His WP48 in the first half was below 0.100.  In the second half is mark has been 0.192.  It is possible that this is related to his position change.  But in general, I have not seen much evidence that changing positions dramatically changes production, although that could have happened in this instance.  It is the case that Mbah a Moute’s number for 2009-10 are about what they were in 2008-09.  So it looks like the Bucks found a productive player in the second round.  The same could also be said about Iyasova.  Again, second round picks can be productive NBA players.  So the Bucks have a chance to add another productive player in the 2010 draft.

Do teams making the playoffs after coming off such a hot run like the Bucks are in the midst of have a better chance of success?  Or will the lack of playoff experience do them in?

I am not convinced playoff experience makes a difference (at least, I haven’t seen any study that suggests this).  Justin Kubatko and I have both won the TrueHoop StatGeek Smackdown with apparently a fairly simple model.  We both appeared to have considered only two factors in making our selections:  The efficiency differential (offensive efficiency minus defensive efficiency) of the two teams and who had home-court advantage.  Playoff experience was not factored into the analysis (at least, I didn’t consider this issue when I won in 2009).

When we look at the 2010 playoffs, it doesn’t look like the Bucks have much chance to go far.  Right now the Bucks would face the Atlanta Hawks in the first round.  It is possible the Bucks could slip a bit (or the Hawks could get a bit better), and that means the Bucks would face the Boston Celtics.  Regardless, though, it doesn’t look like the Bucks would be favored in the first round (although the Celtics are struggling).

How can Milwaukee best use the cap-room they’ll have after the 2010-11 season?  Statistically, is it wise for them to try and obtain a max-level guy (not that that’s necessarily realistic) or would they be better off looking at building depth and facilitating trades for draft picks?

What the numbers suggest is that the Bucks simply have to get more production to advance.  Because the NBA’s labor market is inefficient (again, scoring is overvalued), it is possible to get productive players without acquiring max-level players or high draft picks.

If we look at this team’s roster, it looks like Milwaukee has productive players at every position. What is needed now is some depth.  And it is very possible to find that depth in the 2010 draft (and it is possible to draft poorly and miss out on the few productive players that will be available).  One thing to remember about the draft… productive players can be found throughout the first round and into the second round.  So regardless of where the Bucks pick, it is possible to find help in the draft.

Why should readers be interested in your new book?  Why are your numbers are an effective tool of analysis and just what are wins produced?

About the book ….Stumbling on Wins is a review of decision-making in professional sports.  The now-classic Moneyball story was about how the Oakland A’s managed to win without spending much money.  This was made possible – as Michael Lewis argued (and academic research confirmed) – because Billy Beane knew something about player performance that his fellow general managers had missed.  Specifically, on-base percentage was undervalued.  We told a similar story in The Wages of Wins (our first book) when we noted that scoring in the NBA was overvalued.  In our latest book, Martin Schmidt and I explore a host of similar stories (primarily drawn from academic research published by us and many others) in basketball, baseball, football, and hockey.

What is interesting about all these stories (beyond the fact sports stories are just fun to talk about) is that decision-makers in sports shouldn’t make these mistakes.  Decision-makers in sports are very smart, driven to be successful, have access to an abundance of information, clear incentives, and severe consequences when decisions don’t work out.  Despite these characteristics, though, we see systematic errors throughout professional sports.  Economics, though, traditionally teaches that people do not make systematic errors.  So these stories don’t just tell us something about sports, they also tell us something about how economics should characterize human decision-making.

About Wins Produced… the model used to measure player performance begins with the simple notion that wins in the NBA are determined by a team’s offensive and defensive efficiency.  From this statistical model we can derive the value – in terms of wins – of much that is in the NBA box score.  These values tell us that wins in the NBA are  primarily determined by shooting efficiency and the ability of a team to capture and maintain possession of the ball.  This means that an NBA player helps his team win when he is an efficient scorer, grabs rebounds, captures steals, and avoids turnovers (i.e. the possession factors).  Blocked shots, assists, personal fouls do matter.  But shooting efficiency and the possession factors matter more.

The model explains 95% of team wins and is much more consistent across time than the plus-minus models.  But it does tell us that players who score inefficiently, and/or who are below average with respect to the possession factors, do not help a team win very much.  And this can be true even if the player scores more than 20 points per game.  As a consequence, Wins Produced can produce results that contradict what people generally believe about NBA players.  Again, popular perception is driven by scoring.  So as we argue in the Stumbling on Wins, perhaps what we see from Wins Produced shouldn’t be that surprising.

Let me close by noting that our publisher prefers that everyone know that Stumbling on Wins is now available at and it will soon be available in bookstores.

Categories: 5 Questions

Tags: ,,,,,


  1. Pingback: Five Questions — and Perhaps Some Answers — on the Milwaukee Bucks « The Wages of Wins Journal

  2. Very nice read. This is my first visit to this blog (via Berri’s blog) and I’m not a Bucks fan, but it’s got to be an exciting season for those who are. I think it should be noted that Bogut being healthy played a big part in this season’s success (but he is at the top of that table). Plus Brandon Jennings gave some nice depth at the PG spot, coupled with Ridnour. So a nice young point guard and a nice young center, not to mention the other nice young folks rounding out the rest of the rotation should make Milwaukee a very good team for the foreseeable future. Anyway, thanks for sharing!

  3. Pingback: Daybreak Doppler: You Tell Me « - A Wisconsin Sports Blog

  4. Pingback: Rick Perry, The World, and Stumbling on Wins « The Wages of Wins Journal