By Martin Kleinbard

I’ll start this column with possibly the most self-explanatory table in the history of sports analytics:

NBA Fans 2

The reason for this dramatic shift was, of course, Zydrunas Ilgauskas, who took his talents from Cleveland to South Beach in the 2010 offseason (seriously, look it up). The other player who made that same switch wasn’t bad either.

Which of these two teams has the better fans? It’s an impossible question to answer with just the data that we have here. All we know is that LeBron James’ arrival in Miami caused a huge boost to the Heat’s attendance and a similarly large drop in the Cavs’ (sorry, Big Z), but that shouldn’t come as much of a surprise to any sports fan with a pulse.

What we’re really looking for is a LeBron-independent method of comparing the two fan bases (along with those of the other 28 NBA franchises). Superstars and superteams come and go, and with them so do many fans. As I did in my rankings of the MLB and NFL fan bases, my goal is to control for all of the on-court discrepancies that make being a Thunder fan (excluding this year’s rash of injuries) easier than being a Bucks fan, and in doing so determine the best and worst fans in the NBA.

I used a similar process to that which I followed in my previous two studies: identify the qualities that indicate a good fan base, create models to control for the factors over which fans have no control, and use those models to compare teams’ actual and predicted performance (with the difference representing the fan impact). The time period that I observed spanned the 2008-2009 through 2013-2014 regular seasons.1

What makes good NBA fans? Showing up to games and creating an intimidating atmosphere for visitors are probably on the top of most people’s lists, which is why I used average attendance and home court advantage (as measured by the difference between home and away scoring margin) as my two outcome metrics. I would have loved to have also used local TV ratings (as I did for the NFL rankings), but sadly I was not able to find them in any publicly available source.

As I alluded to in the outset, using raw attendance figures as a proxy for fan quality can be grossly misleading because of the “LeBron Effect.” Thus, I controlled for team performance (wins in the current and previous season) and the number of superstars on a team’s roster (with “superstar” defined as a player who received at least one standard deviation above the average number of All Star votes in a given season).

Since the NBA is often labeled a star-driven league, it’s worth noting that the presence of one superstar is worth the same on yearly attendance as 7 more wins in the previous season or 5.5 more wins in the current season. In other words: stars have a big impact on attendance, but their mere presence can’t move the needle as much as an elite team2. Clearly, this approach oversimplifies a complex issue (LeBron in 2014-2015, for instance, does not have the same value on attendance as Andrew Bynum in 2011-2012, the year he earned enough votes to qualify as a “superstar”), but it’s a solid estimation.

I also controlled for other, non-performance related factors, such as market size and arena quality to not penalize fans of teams in small cities or with shoddy arenas. For the latter, I relied on Yelp ratings (as I did for the MLB and NFL studies).3

Home court advantage is an easier outcome variable to deal with because there are fewer omitted variables. The only one that I ended up needing to control for was the team’s wins in the current season—better teams tended to have higher home-road point splits. Interestingly, attendance didn’t end up having a significant effect—in other words, creating a hostile environment is a function of fan quality, not quantity. Another variable that was not significantly related to home court advantage was team age, which belies the assumption that young teams will struggle more on the road and therefore have higher splits between their home and road performance.

My final decision was how to weigh the two outcome variables. While a formidable home court advantage is certainly the most important action that a fan base can deliver, I don’t have enough trust that the method I’m using to measure it accurately reflects the fans’ “value added”. If I had the average decibel level in each arena, that would be a different story. Without such a stat, I’ll weigh attendance at 70% and home court advantage at 30%.

Without further ado, here are the rankings, as measured by standard deviations above or below average:

NBA Fans

As if winning the World Series last month wasn’t enough good news for the Bay Area, the honor of best fans in the NBA belongs to “Dub Nation” —aka the fans of the Golden State Warriors—with a weighted average of 1.09 standard deviations above average (SDAA). Despite playing in one of the worst arenas in the league and fielding a team that, until 2012, was a perennial loser, the Dubs sold out nearly every game. Nipping closely at their heels are the plaid-adorning fans of Portland Trail Blazers at an even 1.00 SDAA, comfortably above the fans of the bronze medalist Utah Jazz (0.80).

At the bottom of the spectrum lies the City of Brotherly Love, where the Philadelphia 76ers sit 1.35 standard deviations below average. Before you go up in arms about the team’s recent tanking clouding the picture, remember that the rankings take records in the current and prior year into account, and that the Sixers actually had three seasons at or above .500 during the time period studied.

A couple of teams warrant special recognition for the disparity between their attendance and home court advantage rankings. The Sacramento Kings are a pedestrian 13th in attendance but 1st—by a long shot—in home court advantage. Apparently, selling the arena name to a company called Sleep Train didn’t have a drowsy effect on fan noise. On the flip side are the Chicago Bulls, who rank 5th in attendance but 28th in home court advantage. Maybe they need the team’s epic 90s intro montage to fire the crowd up.

As was the case with the NFL rankings, this approach is a bit unfair to teams like the Lakers that sell out almost every game but are predicted to do so because of the team’s success (again, excepting this year’s horrible start), market size, and star power. The Patriots are the obvious football parallel. Thus, their maximum possible attendance “value added” is limited in comparison with a team like the aforementioned Warriors. That being said, LA’s home court advantage ranking of 21st certainly didn’t do much to shed its fans’ reputation for being more interested in watching celebrities in the stands than the action on the court. The only other team that could be characterized as victim of its own success was the Spurs, but they too had a pedestrian home court advantage figure and thus have a minimal gripe. In short, teams toward the middle of the rankings may be mildly mis-seeded, but those at the top and bottom almost definitely deserve to be there.

I’ll end with a quick summary of my three sports’ fan rankings. There are 15 cities with at least one team in each of the “Big 3” leagues (counting New York/New Jersey and the San Francisco Bay Area as one “city” each). Taking a simple average of each city’s individual team rankings, I came up with a very crude method for determining which metropolis has the best fans in the country. Boston? Chicago? The Big Apple? Try the Mini Apple—as in Minneapolis—which rode a dominant #1 ranking in baseball to the top. Considering that it finished #11 in football and #22 in basketball, I’m not yet ready crown Minny king of American sports.

I am, however, ready to adorn Miami the distinction of worst sports city in America. That’s what happen when you finish dead last in baseball and football and fifth to last in basketball; not even LeBron could help it avoid that fate. Then again, given Miamians’ other entertainment options, I doubt that they’re too hung up about this title.

***

Martin Kleinbard is the co-founder of The Bandwagn, a newsletter/website that allows non-sports diehards the chance to join the sports conversation. He can be reached at martin.kleinbard@gmail.com.

 

1. 2008 was a convenient year at which to start because it was the Thunder’s first season in OKC and there have not been any significant relocations since then (the Nets moved from New Jersey to Brooklyn in 2012 but at least that’s the same metropolitan area). In addition, that’s also enough time to cycle through good and bad times for most teams: only three teams didn’t have at least one winning season in this time period, and only five didn’t have at least one losing season.


2. For those concerned about multicollinearity (i.e. the fact that teams with superstars also tend to be really good, which clouds the ability to compare the two variables), note that the correlations between superstars and wins in the prior season and this season are 0.42 and 0.50, respectively: large, but not large enough to invalidate using both. In addition the VIF of all 3 variables are all 2.08 or lower, indicating a minimal multicollinearity risk.


3. One clear shortcoming of this method is that Yelp does not force users to distinguish between their thoughts of the arena for basketball use as opposed to hockey, concerts, and other performances held in those venues. That being said, since most aspects of the arena (location, food options, seat comfort) are the same across entertainment type, I’m not overly worried about the multi-use issue throwing off the rankings. In case you’re wondering, Memphis and Indiana hold the top two spots of the arena rankings, while Minnesota and Sacramento bring up the rear.

You May Also Be Interested In These Stories

[display-posts category=”sports” wrapper=”ul” posts_per_page=”4″]