The Standard Small Sample Warning!

Once upon a time, I wrote virtually everything that appeared at the Wages of Wins Journal Journal.  During the NBA season – which lasts from late October to June (or much of the year) – posts followed this format: Data would be collected and analyzed for a specific team. With the team analyzed (i.e. Wins Produced calculated), the story the data suggests would be told.   Since I was the only person calculating Wins Produced, I had a monopoly on the stories this approach offered.

A few years ago, though, Patrick Minton figured out how to automate the Wins Produced calculation.  The result was the theNBAgeek.com.  Patrick’s website provided every team’s Wins Produced throughout the season.  With this data in hand, my monopoly was eliminated.  Anyone interested in the Wins Produced story could simply look up the data at theNBAgeek.com.  Consequently, I asked for other people to start telling the stories.

A number of people answered the call. But a trio – Andres Alvarez, Arturo Galletti, and Devin Dignam – wrote most of the new stories at the Wages of Wins Journal.  Unfortunately, Andres, Arturo, and Devin have now joined forces with Patrick.  Their new website – boxscoregeeks.com – provides data, graphics, and analysis.  In sum, these guys are doing way more than I ever hoped to do.

So what can we do at the Wages of Wins Journal? Some people have volunteered to start writing in this forum.  Hopefully their work will start appearing soon.

For now – as I noted a few days ago — we are back to me doing the writing.  My plan is to go back to the model I followed a few years ago.  In other words, I hope to write posts that primarily focus on a specific team.   This should be easier, since Patrick and company are going to provide the data analysis. But I am now facing serious competition.  Hopefully, though, I can come up with some thoughts worthy of your attention.

Before I get started, though, let me devote this post to an important — and very standard early-season — disclaimer.

At this point, teams have played six or fewer games.  Imagine if these games had occurred in February.  If you looked at six games in February would you be inclined to think that those games could tell you which players are among the “best” and “worst” in the NBA?  Hopefully you recognize that small samples – whether the sample is in February or November – are not very informative.

So entering tonight’s games, we see (among a few other names with 100 minutes played) that Paul George, Klay Thompson and C.J. Miles have WP48 marks beyond 0.300.  Last season, only Chris Paul, LeBron James, and Kevin Durant finished the season with WP48 marks above 0.300 (and with 2,000 minutes played).  Although it is possible that George, Thompson, and Miles are now on par with the top three players in the game today; a more likely scenario is that these players have just had a few games where they played well.  And soon, that will not be the case.

Likewise, Derrick Rose is currently the least productive player in the game. Rose is certainly a bit overrated by the media.  But it seems very unlikely that Rose is really among the worst players in the game.

So although I will soon be writing about these small samples (because they are only samples we got!), we need to be sure we don’t take these samples too seriously.  It really is going to be a few weeks before we can start to make sense of the 2013-14 season.

And that means we need to keep this disclaimer in mind: Just because you see some numbers right now, it does not mean those numbers can tell us much.  Or as I tell my students, sometimes the answer is “I don’t know.”  We are often reluctant to say this.  But when the samples are small, that is often the best we can do.

– DJ

P.S. Again, though, I am going to tell stories with these small samples across the next few weeks.  But I just wanted to be clear, nothing definitive can really be said right now.  However, though, a bunch of non-definitive statements can certainly be offered!

Comments are closed.