The NBA regular season starts tomorrow, which means it’s time to put out projections for how many wins I think each team will get. I tend to approach an exercise like this using numbers instead of my subjective opinion. Why? Because my subjective opinion has no game.
One problem: there’s no shortage of statisticians among NBA prognosticators. So instead of coming up with some genius system that tries to compete with everybody else, I’ve decided to be as minimalist as possible. With that in mind, I’m using a grand total of two data points for my analysis: a player’s minutes projection (straight from ESPN) and his regularized adjusted plus-minus rating. I go into this in a little more detail here.
The point of this is not to be the single most accurate person in making predictions, but instead to see just how well I can do with the least amount of information or subjectivity.
Here are the results:
If a team’s projection looks higher or lower than it should, it’s probably because of the plus-minus ratings. Here are some specific examples:
I don’t think anybody else has the Clippers as a championship contender this year. But Chris Paul (+6.0) and Blake Griffin (+3.7) are a fantastic duo, and they will hopefully get a full year out of Chauncey Billups (+1.3).
The Hawks have Josh Smith (+4.1), a healthy Al Horford (+1.3), Louis Williams (+0.9), Jeff Teague (0.0), and Devin Harris (+0.5). That’s not enough to be an elite team, but I don’t think 52 wins is out of the question.
Many have Dallas finishing out of the playoffs since Dirk Nowitzki (+8.0) figures to miss a significant number of games. But my method keeps Dallas in the playoffs because of Nowitzki, who had the NBA’s highest adjusted plus-minus rating last year.
Deron Williams (+2.4) is good, but something less than a superstar. The same could be said of Gerald Wallace (+2.1) and Joe Johnson (+1.5). Unfortunately, the Nets are paying them like superstars. What kills them is their bench, which mostly consists of guys who can score but are rated horribly by adjusted plus-minus.
Paul Millsap (+5.0) is outstanding. But after Millsap, the highest-rated players on the Jazz are Earl Watson (+0.8) and Al Jefferson (+0.1). Quite a few of their players are big negatives, particularly Enes Kanter (-4.2).
A lot of people think the Warriors could make the playoffs this year. I felt that way too, but I count exactly two players who rate as positives for Golden State: the injury-prone Stephen Curry (+4.0) and the injury-prone Andrew Bogut (+2.8). I get the optimism, but the potential for disaster is there as well.
THE HARDEN TRADE
For those curious about the impact of the James Harden trade, I previously had Oklahoma City projected for 57 wins, and Houston projected for 29, so the Thunder lost seven wins and the Rockets gained six. I gave the spare win to Utah, the closest team to being rounded up (32 to 33).
Thanks for checking out my projections, and here’s hoping that I neither finish first (because that would say horrible things about the NBA stats community) or last (because I don’t want to finish last).