How NBA Twitter fixed basketball’s bad officiating

Increasingly sophisticated data and vocal fans helped unrig the system

Happy Juneteenth! It’s a federal holiday, so we’re off today, but enjoy our intern Maya’s first Slow Boring post.

There’s little NBA fans seem to enjoy more than complaining about the referees.

These concerns are easy to dismiss as conspiracy theories, especially when they come from the opposing team’s fans, but there is strong evidence that referees are motivated by psychological factors to favor (1) the home team, (2) the team that’s behind in a game or playoff series, and (3) superstar players.

Until recently, the NBA tolerated — or maybe even subtly encouraged — these biased calls because they were good for business. But this all changed with the rise of digital and social media. Today, fans, journalists, and academics can easily access game data and hold the NBA accountable online. The massive NBA Twitter community has become the ultimate check on bad refereeing and has forced the league to crack down on biased officiating.

Favoring the home team

In the NBA and professional sports in general, the home team has traditionally had a huge advantage:¹

(Source: Scorecasting)

Looking just at the NBA, in the seasons from 1999-2009, home teams won 60% of regular season games and 65% of playoff games. In their 2011 book “Scorecasting,” Yale behavioral economist Tobias Moskowitz and “Sports Illustrated” editor Jon Wertheim identified referee bias toward the home team as the primary factor in generating this home-court advantage, largely due to three distinct disparities:

• More free throws. NBA referees awarded the home team one to 1.5 more free throws per game.

• More favorable discretionary turnover calls. This was quantified in a 2012 study by BYU Professor Joseph Price, who found that referees granted the home team 11% more discretionary turnovers per game in the same period.

• More aggressive play. The expectation of lenient officiating encourages home teams to play more aggressively, and the expectation of more harsh officiating toward visitors encourages them to play more carefully.

Moskowitz and Wertheim found that this combination of disparities gave home teams an average of over 2.5 points per game and accounted for three-quarters of the home-court advantage in the 2000s. 

They also show that other commonly cited explanations — weariness from travel, unique arena characteristics that the home team is comfortable with, and unfavorable scheduling for the visitors — fall short. Visiting teams may be tired from travel, but home teams beat visiting teams at the same rate even when they are based in the same city and travel is marginal. And in the NBA, home teams don’t benefit significantly from unique arena characteristics because arenas are mostly standardized.

Moskowitz and Wertheim do agree that scheduling is an important contributing factor to home-court advantage — road teams play more tiring back-to-back games than home teams do. But the authors conclude that this only accounted for one-quarter of the home team advantage in the 2000s.

The data is clear: referee bias is the number one cause of the home-court advantage.

Other referee biases

In addition to favoring the home team, referees are biased toward the team that is losing, both in regular season games and in the playoffs. From 2002-2008, referees called 15% fewer discretionary turnovers on teams trailing by more than 10 points compared to teams that are trailing or winning by less than three. On the flip side, referees called 10% more discretionary turnovers on teams winning by four to 10 points.

This tendency has not escaped the notice of fans, who regularly complain about referees like Scott Foster (nicknamed “The Extender”), and certain games (like game 6 of the 2002 Western Conference Finals) are particularly controversial. And it’s not just about individual games — fan conspiracy theories were encouraged when disgraced former referee Tim Donaghy, who was sentenced to prison time for betting on games, alleged that the NBA fixes games to extend playoff series. 

Looking at the data, it’s clear that referees do make calls that favor the team losing the series. Between 2002-2008, for each game a team was down in the series, referees called 3.4% more discretionary turnovers in favor of that team.²

But this is only a small advantage. Because of the way playoff series are structured, home-court bias actually plays a much bigger role in extending playoff series, as economist Kevin Hassett has shown.

For most of NBA history, all playoff games were played, as they currently are, in the 2-2-1-1-1 format. If the visiting team is down 2-0, 2-1, or 3-2, the home team ref bias can help them catch up in games 3, 4, and 6.

From 1985-2013, the structure was slightly different and even more amenable to the home-court advantage extending series. All playoff games were played in the 2-2-1-1-1 format except for the finals, which were played in a 2-3-2 format. According to a paper by 1998 paper by FAU Professor Steven Caudill:

“…the HHAAAHH format leads to a longer series. The Second team gets to play three of the first five games at home. With a home court advantage, the Second team is more likely to win those games and the series is likely to continue.”

The old finals format allowed the lower-seeded team to play three out of five of the first games at home, compared to two out of five in the new format. Giving the weaker team the home-court advantage at the beginning of the series made it less likely that they would be eliminated early. As a result, longer series were more likely.

Also worth noting: home-court advantage decreases significantly in game 7. Once a series has reached its final game, referee bias for the home team could decline because the teams are tied and the underdog bias no longer applies. 

Referees also appear to give superstar players more favorable calls.

Fans regularly circulate videos that show referees allegedly giving special treatment to the league’s best players, and there’s more than just anecdotal evidence for this. Multiple analyses of games from 2015-2019 found that the NBA’s biggest stars benefit from referee bias in two ways. First, superstars are disproportionately likely to receive incorrect “no-calls” where they commit a defensive foul but do not draw a call, which makes it less likely that they foul out of games. Second, superstars draw more favorable fouls than other players. Referees call a disproportionate amount of fouls against the players defending stars, sending them to the free throw line to shoot more. 

Superstar bias becomes even more consequential in playoff games. During the fourth quarters of the 2011 playoff games, referees gave All-Stars an average of 0.32 extra free throw attempts per minute. If a star plays for all 12 minutes of the final quarter, that amounts to almost 4 additional free throw points on average — that could make or break a close game!

The psychology behind biased officiating

Referee psychology is the simplest (and least sinister) explanation for bias. Social pressure matters! Hometown crowds often scream at the referees when they make calls against the home team. Forced to make split-second decisions, referees are (perhaps unconsciously?) impacted by this pressure and make more calls in favor of the home team.³

The COVID-19 pandemic created a natural experiment to assess the role of the crowd because many games in the 2020-2021 season were played in empty arenas.

As you can see in this graph from FiveThirtyEight, home teams gained an average of almost two points per game when they played in front of crowds versus when they played in empty arenas. This could be due to the crowd impacting the officials or the players. However, comparing free throw percentages at home vs. away helps separate these factors. If crowd cheers really encouraged the home team and boos really got to the heads of the visitors, we’d expect that teams would make more free throws at home. But during the 2020-2021 season, teams made free throws at very similar rates at home vs. away. In fact, many teams even shot better when they weren’t at home! It certainly seems that hometown crowd pressure does not have a huge impact on the players and that the crowd’s impact on referees is the true source of the home-court advantage.

With regard to trailing team bias, psychologists have found that humans tend to empathize with and root for underdogs. Even if referees try to maintain impartiality, they may be subconsciously influenced by this bias and call more fouls for the trailing team. In the playoffs, referees face heightened social pressure from the crowd (with more fans watching) and heightened underdog bias (because each game has more impact), which can have the effect of extending the series. 

Finally, superstar bias can be explained by referees admiring LeBron and Steph, just like normal NBA fans. Under stressful, quick decision-making conditions, these biases can seep into decision-making — nobody wants a famous superstar to get mad and argue with them on camera! Furthermore, as with home bias, crowds can pressure referees to give lenient calls to hometown stars. The evidence backs this up: an 1987 study found that NBA referees called fewer fouls on star players during home games in the 1980s.

The NBA allowed this to happen

Even though there are strong psychological explanations for referee bias, these biases are also very profitable for the NBA — maybe coincidentally, maybe not. If home teams win at a high rate, more home fans will enjoy games. If fans expect the home team to win, they will buy more tickets. It was financially good for the Warriors this past season that they were so much more successful at home (33-8) than on the road (11-30); all things being equal, teams would prefer their wins to come in front of their own fans. 

It’s also in the NBA’s interest for games to be close, because more people will keep watching and tune in late to see how a nail-biter ends. Longer playoff series allow the NBA to sell tickets and profit from television viewership (which increases as the series continues) from more games. Take game 7 of the 2016 NBA Finals — the last time there was a Finals game 7 — as an example of how lucrative long playoff series are for the NBA. This game was the third most-watched game in NBA history with an average of 31 million viewers throughout the game. Note that the first and second most watched games were also late-stage games: games 6 in 1998 and 1993. Late-stage games are exciting, so exciting that the league was able to charge $122,000 for court-side seats and $1,000 for the average ticket for game 7 of the 2016 Finals. This game clearly helped the league make enormous amounts of money — and that’s compared to the $0 they would have made if the Warriors beat the Cavaliers in game 6 and there was no game 7.

Finally, favorable calls for superstars also help the NBA increase profits. The NBA relies heavily on the fame and marketability of its superstar players. The most exciting periods in NBA history have revolved around superstars: Magic and Bird in the 1980s, Michael Jordan in the 1990s, Kobe and Shaq in the 2000s, and Steph and LeBron in the 2010s. The NBA wants its superstars to shine. Referee leniency towards superstars on offense allows them to score more points, create more exciting plays, and remain in the game for longer (i.e., not be subbed out due to foul trouble). Calling “phantom” fouls on the players defending stars also helps drive up the superstars’ points. Higher-scoring games (especially for star players) with lots of superstar playing time and exciting, acrobatic plays often translate into higher TV ratings.

Is this all just a happy coincidence?

I am not a conspiracy theorist and I think it’s highly unlikely that the NBA explicitly ordered referees to make biased calls to enhance revenue. If the NBA tried something nefarious along these lines, it’s almost certain that someone (a disgruntled ref or league exec) would leak it — or it would come out in litigation.

That said, it’s pretty convenient that referees' natural biases align so neatly with NBA financial interests, and it’s possible that the NBA subtly incentivizes referees to indulge these biases. 

This is difficult to prove because the NBA would have every incentive to conceal such an initiative. However, there is some supporting evidence. Price has pointed out that the NBA monitors referees very intensely. If the league wanted to incentivize certain biases, they have the information necessary to do so. Price found some evidence that this may be occurring (emphasis mine): 

We find that refs who work more playoff games are more likely to be assigned to regular season games involving weak home teams playing strong opponents. Because these are the games in which the benefits to the league of home favoritism would be relatively high, this indicates the league makes ref assignments in a strategic way, and rewards refs who help home teams when they need it most.

The NBA may assign certain referees who often help extend games (like Scott Foster) to crucial playoff games. Price argues that the NBA rewards referees who exhibit more home-court bias by giving them the lucrative and prestigious opportunity to call playoff games. This both encourages home court bias during the regular season and helps ensure that this bias persists in the playoffs, providing a boost to team and league finances.

The Internet forced the league to mitigate referee bias

Perhaps surprisingly, given the financial incentives, referee bias has significantly improved in recent years. A new study suggests that referee home bias has significantly decreased (although some of this was due to the effects of empty stadiums during the pandemic). While 65% of home teams won games in the playoffs during 1998-2008, in recent seasons this advantage has been decreasing and is now below 60%.⁴ And last week, Bill Simmons pointed out that during 2018-2023, excluding the 2020 Bubble, the road team has won more finals games than the home team.

So what changed? I think the rise of the internet, and NBA Twitter in particular, has shifted the league’s business calculus, democratizing access to real-time NBA data and increasing scrutiny on officials.

Before the Internet, it was hard to even figure out which referees had officiated a game after the fact. Today, anyone can access a huge amount of data about player statistics, officiating, and fouls in any NBA game. Fueled by this data, NBA Twitter provides an international social media platform for fans, journalists, and academics to discuss every aspect of the game — and there’s no topic NBA Twitter loves more than biased officiating. More advanced camera angles and online videos allow fans to post and discuss specific clips of bad calls (and even referees’ facial reactions), which regularly go viral with tens of thousands of users lambasting bad calls to an audience of millions. And the harsher and more critical the take, the more viral the distribution.

But while referee bias might boost short-term fan enthusiasm, repeated evidence of bias could cause fans to distrust the league and watch fewer games. It’s in the long-term interest of the NBA to be perceived as fair.

In recent years, the NBA has taken three key steps to combat referee bias:

• In 2015, the NBA started releasing Last Two Minute Reports which analyze calls made in the final minutes of close NBA games and highlight incorrect calls and no-calls.

• The league also introduced the Coach’s Challenge in 2019. This allows a coach to ask to review one call per game. Both the Coach’s Challenge and the Last Two Minute Reports lead to the possibility that a referee’s call is labeled as incorrect.

• The league has taken advantage of its increased access to data to start disciplining referees for bad or missed calls.

These changes are direct disincentives for biased officiating. In addition to proactive changes from the league, online scrutiny of referees may have helped reduce bias on its own via direct social pressure — refs see viral tweets, too.

Something similar happened 15 years ago. In 2007, Joseph Price⁵ published a study on racial bias in NBA refereeing. Since then, racial bias in refereeing has essentially disappeared. The NBA claimed that it hadn’t made any policy changes related to racism bias — it’s possible the refs self-regulated and changed their behavior on their own. It’s also possible that the same effect occurred in recent years with increasing awareness of other types of referee bias.

As Supreme Court Justice Louis Brandeis said, “Sunlight is said to be the best of disinfectants.” By shedding light on referee bias, digital media and NBA Twitter have forced the NBA to solve a decades-old problem. Finally, the league is cracking down on biased officiating and trying to play more fair. 

(Thank you to Randy Cohen for his help with this post.)

1. There are also many studies about referees favoring the home team in other sports, especially European soccer, that support this argument.

2. Multiple studies have highlighted a similar effect in college basketball.

3. In a 2002 study, researchers showed referees videos of soccer tackles, some with crowd noise and some without. The referees that heard the crowd noise were 15.5% more likely to call the tackles for the home team (which the crowd was cheering for) than those who did not hear the noise. While this evidence is from another sport, it shows the general psychological tendency of referees to be biased by the crowd (and therefore, in favor of the home team).

4. 59% of home teams won in the 2023 playoffs, 60% won in the 2022 playoffs, and 56% of home teams won in the 2019 playoffs. I’m ignoring the 2020 Bubble playoffs.

5. Price is the same author who wrote “Sub-Perfect Game: Profitable Biases of NBA Referees,” another major study I reference throughout this article.