ESPN has been running a series on each NBA team recently, asking five writers five questions about each team. Today was my team, the Pistons. The last question, a tough one for Pistons fans, was when we can expect to see Detroit in the playoffs again. There were three votes for 2013, one for this year (if it happens), and one for 2017. 2017! The Pistons have already missed the playoffs for two seasons; if the writer optimistically meant 2016-2017, that would be eight years in a row. That seemed almost too long to be true. I was curious; how long can you really expect a team to stay out of the playoffs?
On the one hand, the NBA is not a place where teams turn around their fortunes. Typically good teams stay good and bad teams stay bad. On the other hand, that’s only ‘typically’, and (at least currently) in the East you can make the playoffs even as a fairly poor team. To look into things, I got the basketball-reference.com team page for each current NBA team (here’s the one for the Pistons). For each season for a team, I recorded how long it would be for the team to make the playoffs. For example, in 2000-2001 the Pistons were one year away because they made the playoffs in 2001-2002. Any year a team made the playoffs they got a 0, and teams that are currently on a no-playoffs streak got an NA (which would include the Pistons) because we don’t know when they’ll make the playoffs next.
To predict when a team will make the playoffs again, I first removed the current no-playoff streaks and the first seasons for each franchise (because they have no previous season, which I’ll use later). I ran a generalized linear regression (using the Poisson distribution; technically, something a little fancier even but that’ll do) predicting years to playoffs from a few different measures. I used the number of years last season and either the current season’s win percentage, SRS, or offensive rating minus defensive rating. Rating differential couldn’t be compared to the other two directly because rating isn’t available for some seasons, but win percentage actually did better than SRS (perhaps because wins decide who makes the playoffs, regardless of measures that might be better indicators of team quality); to the best I can tell, win percentage and rating difference do about equally well. What’s interesting is that whether you use win percentage or rating differential, the model does better when you include the square as well. You can see why if you look at the plot of win percentage against years to the playoffs, or the model prediction.
The actual data is in the first plot; you can see that the longest any team ever had to wait to get to the playoffs is ten years, which was the Mavericks. Keep in mind that the data are cumulative; the year after the ten year wait, the Mavs still had to wait nine seasons, so they contribute one of the two nine-year points (the other belongs to the New Orleans Jazz). You might notice a bulge out to the side toward the middle of the plot; that’s the effect of the squared term. It’s more obvious in the second plot, which is the model prediction using win percentage and percentage squared. If a team’s win percentage is 50% or better, they’re likely in the playoffs and thus have a 0 season wait. Similarly, if a team is really bad they have a short wait, probably due to getting a great draft pick. But, if your team is only sort of bad, they can have a wait of up to 5 seasons (the full model, which includes the previous season’s wait and finds that it’s bad news, predicts up to 13 seasons).
This fits with an argument that a number of people have made about winning championships, which is that you want to either be really good or really bad. If you’re only kind of bad, you don’t get good draft picks and you aren’t good enough to benefit from signing a good free agent (unless he’s really, really good).
So, when can the Pistons (and other teams who missed on the playoffs this year) expect to be back? I used the win percentage model (including the lagged wait) to predict how long they have to go after this season. In alphabetical order: Charlotte (made it last season) has just over a season to wait, meaning they have a good shot at making it next year (if there is a next year). Cleveland (made it last season) has closer to two seasons to wait. They might not be fit very well given the circumstances of how they made it before, namely by having LeBron, but they did just get two high draft picks. Detroit is probably a couple seasons away, maybe 2012-2013. Golden State hasn’t made the playoffs in four years, which is a bad sign; they’re also probably a couple seasons away. The Houston Rockets were actually an above-500 team, so they are expected to get there next year. The Clippers, who some people are eying as a playoff team, are probably 3 seasons away. They’re right in that unfortunate ‘mediocre’ spot. The Milwaukee Bucks might make it this year, while the Minnesota Timberwolves and New Jersey Nets probably have a few years to go. The Suns and Jazz may or may not make it while the Kings, Raptors, and Wizards don’t have a chance.
So if you had to pick the teams to get back in this season, you might want to go with Charlotte, Houston, Milwaukee, Phoenix, and Utah. Who would drop out? Probably the teams towards the bottom of the playoffs last year, like Indiana, Philly, Memphis, and New Orleans. Again, you want to be really good or really bad; there’s no use in being in between.