Movies

Moneyballing the Movies: How the Box Office Became a Sport

Every weekend, new releases are graded on their box office returns. How did covering the movie business become a haven for statheads?

Movies

Moneyballing the Movies: How the Box Office Became a Sport

Every weekend, new releases are graded on their box office returns. How did covering the movie business become a haven for statheads?

By Shaun RavivAug. 2, 2018, 10:30 am UTC • 16 min

What is the biggest movie of all time? Easy. It’s Star Wars: The Force Awakens. America rules, and The Force Awakens grossed $936.7 million at the U.S. box office after its release in 2015, by far the most ever. Next on the list is the $760.5 million made by Avatar, released back in 2009, then Black Panther at $700.0 million (and counting). No contest. Hand the crown to Rey and Finn and bow before them.

Who is the greatest player in NBA history? Easy. It’s Kareem Abdul-Jabbar. He scored 38,387 points in his career, nearly 1,500 more than second-place Karl Malone, and no current player is anywhere close to catching him.

But anyone who pays attention knows the greatest player of all time is constantly up for debate. The parameters are endless: Who has the most rings? Who played elite defense? Who won the most MVPs? Who was durable? Who made their teammates better? Who was their competition? Were they playing against guys who dribbled with only one hand and shot underhand free throws? Did they play against only white people? In The Book of Basketball, Bill Simmons built the template for an ever-evolving pyramid of NBA greats, and he’s one of many sportswriters who have spent their careers trying to place past and present athletes on the same court.

The same goes for the box office. The Force Awakens was a cinematic spectacle, released after fans of the original trilogy waited three decades for a decent Star Wars movie. But despite sitting atop the domestic box office dollars list on both Box Office Mojo and The Numbers, the two most complete online compilers of box office data, The Force Awakens is not the biggest movie of all time. Which movie is? The answer isn’t so easy to come by.

Reading the standard “who won the weekend” reports now found in nearly every major American newspaper will leave you, at best, modestly informed about the true box office winners and losers. Since I first opened an issue of Entertainment Weekly in 1994, box office statistics have advanced along with sports statistics, moving well beyond the routine box score. Following the box office is similar to following sports in a hundred ways, and it took me more than 20 years to figure out that that’s why people like me, with no skin in the Hollywood game, obsessively pore over the numbers.

Bruce Nash always followed the money. The former IBM software developer founded The Numbers in 1997 after moving to Silicon Valley during the dot-com boom. A sports stats nerd and movie fan, Nash thought he’d try to combine his hobbies into a business, so he began compiling and organizing box office numbers into coherent lists. He’s been doing it ever since. Nash knows that the question of the biggest movie of all time is a complicated one.

Titanic opened in December 1997, about two months after Nash started The Numbers, so the film is dear to his heart. And it’s in the running for the biggest box office hit of all time. Titanic is a quality movie creatively, he says, but it’s difficult to separate that quality from the sensation it created, filling theaters for months after its release. It was the perfect storm for The Numbers. “Here comes Titanic and it’s just breaking all of these records and there was so much interest around that,” he says. “It made me feel that I was doing the right thing with the site.”

But nostalgia aside, Nash—who also works with Guinness adjudicating obscure box office records—says that despite 20 years in the movie numbers business, identifying the biggest box office hit is something he’s still trying to do.

There are three major issues: (1) 80-year-old box office smashes like Gone With the Wind and Snow White and the Seven Dwarfs have been rereleased in nearly every decade since the 1930s, which makes adjusting for inflation a challenge; (2) international box office numbers have been growing rapidly, putting past movies at a great disadvantage compared to modern hits; and (3) domestic theaters don’t release the actual number of tickets sold for each film, only the gross box office dollars earned.

Gone With the Wind is at the top of Box Office Mojo’s Domestic Grosses—Adjusted for Ticket Price Inflation list, but it’s not on The Numbers’ list. Nash says they’re still working to create what he calls “the definitive list” of inflation-adjusted films, so for now they include only films released in 1977 or later.

Comparing Gone With the Wind to Titanic to Black Panther is like comparing Wilt, Jordan, and LeBron. They played in different eras against different levels of competition. When Wilt played, there was no 3-pointer. When Jordan played, there was no hand-check rule or zone defense. LeBron plays in a time when three of the greatest shooters in history are on the same team and have the benefit of advanced offensive and defensive schemes. These arguments are too often reduced to raw numbers, emotion, and personal bias.

Titanic’s original run ended in about nine months, lengthy by today’s standards. Thor: Ragnarok, a massive Marvel hit, was out of theaters in four and a half months. Gone With the Wind’s original release (actually three releases back to back to back) lasted four years. And back then, box office data wasn’t collected in the same way it is today. One estimate put the number of Gone With the Wind tickets sold, just during those early runs, at 60 million, equivalent to nearly half the population of the United States in 1939 — but with subsequent releases that number rises to about 200 million tickets. Mojo estimates Gone With the Wind’s domestic take at $1.9 billion, nearly double The Force Awakens’ unadjusted (or adjusted) take.

Although Black Panther quickly became a cultural phenomenon and has now made more money than Titanic at the box office, its run will probably come to an end about five months after its release. It has sold in the neighborhood of 75 million tickets. The Wakandans are ranked third on the all-time domestic list (just ahead of Avengers: Infinity War, which probably won’t close the gap), but only 30th on Mojo’s all-time domestic adjusted for inflation list.

When Dirk Nowitzki emerged in the NBA as a superstar with an unprecedented jumper for a 7-footer, it heralded a new era. Every team wanted to find the next Dirk, and the systematic pool for recruits increased from the U.S. to Earth. Every draft’s first round since has featured at least one player born outside the U.S., and today it’s simply accepted that the NBA is a world league.

A similar thing has happened at the box office, but to a far greater extreme. While domestic sales remained essentially flat from 2011 to 2017 at about $10 billion to $11 billion, the international market has grown from $22.4 billion to $29.5 billion in that same time. The biggest domestic grosser last year wasn’t Star Wars: The Last Jedi at $620 million in the U.S., but Wolf Warrior 2, which took in $854.2 million in China. Which is to say that if you’re going to calculate the biggest hits today, you can’t ignore international theaters. Without accounting for inflation, the international champ is undoubtedly Avatar, which took in more than $2.7 billion total riding its 3D novelty, $500 million more than its nearest competitor.

Along with Gone With the Wind, Titanic, and Avatar, Nash also puts Doctor Zhivago and Lawrence of Arabia in the running for the biggest box office hits of all time. Though he’s skeptical of some early box office reporting, he and his team at The Numbers scour old newspapers to find box office data from the early and mid-20th century. The Numbers and Mojo are essentially box office versions of Baseball-Reference, which relies on data from the early baseball tomes, which in turn relied on old newspapers and stathead fanatics to populate their historical data. But not all Babe Ruths are created equal.

In 1994, my older brother subscribed to Entertainment Weekly. Only a high school freshman at the time, I’d get to see the issue after he was finished or, if I was lucky, sneak in a look before he noticed it had arrived in the mailbox. I’d read the articles, but what I was really interested in were the lists, specifically the weekend box office numbers.

I remember literally shouting for joy when The Crow opened with a whopping $11.8 million over the weekend, besting all in its path. Of course it was almost two weeks after the fact when I found out The Crow had won because EW was generally two weeks behind in its reporting. (Not bad in the era before high-speed internet.) When the numbers came in, I’d already seen The Crow, thought Brandon Lee was a badass, and wanted him to rule the box office from beyond the grave.

To take the crown that weekend, The Crow had bested the ho-hum openings of When a Man Loves a Woman and Crooklyn, the third weekend of With Honors, the ninth weekend of Four Weddings and a Funeral, plus nonopening weekends of 3 Ninjas Kick Back, No Escape, and Clean Slate, none of which any human being has ever seen. It wasn’t exactly a midsummer blockbuster weekend. This was my first lesson in box office relativism.

Today, everywhere from BuzzFeed to Forbes, The New York Times to the Hattiesburg American publishes box office numbers. It’s nearly as routine as checking the box scores. And EW didn’t originate box office reporting, but rather only mainstreamed it.

Variety has been covering box office grosses for nearly a century. On January 7, 1925, the paper reported that the silent film Peter Pan had brought in $59,503 from two theaters in New York, dominating the Broadway cinemas, which combined had earned a total of $250,000 the previous week. By the 1940s, Variety was covering the box office in 25 cities, with numbers from each collected by a different local stringer. According to a 2003 article by the late film section editor Herb Golden, this is how it worked back in the day:

I learned the hard way about the accuracy of b.o. numbers from my three years as Philly stringer for Variety. I was one of the 24 muggs who had to provide the weekly grosses in his town. You called all the house managers on Monday to get the weekend figures and some were cooperative and some were not—and some you often couldn’t reach. But the story had to be in NY on Tuesday morning, so there was obviously fudging, estimates from competing house managers, local sales managers of the film companies and any other source you could develop.

Golden claims to have “initiated the current plethora of national boxoffice numbers” by looking at the results from all 25 cities together and declaring which films were doing well and which not so well. He also created the first annual compilation of highest-grossing films, and indeed in the January 1950 issue you can see the list of 1949’s biggest films, topped by Jolson Sings Again at $5.5 million. That reported figure was not the gross box office earnings, the number generally reported today, but the film “rentals,” the cut portioned by the theater to the distributor, which has varied over time and today varies even within a film’s run.

The weekly haphazard regional analysis continued into the 1970s, according to Open Wide: How Hollywood Box Office Became a National Obsession, by Dade Hayes and Jonathan Bing. They write that on June 25, 1975, a few days after Jaws opened, Variety ran 13 stories about its grosses in different U.S. cities. A Variety writer named Art Murphy was a pioneer in bringing the national data together each week, eventually turning his work into Art Murphy’s Box Office Register, an annual compendium of box office data, which included grosses and per-screen averages, among other metrics. The Register ended in 1995, but by that time box office grosses were a weekly national news story in EW and elsewhere.

Nowadays, nearly all the box office data that you see comes via comScore, the massive media measurement company—which gobbled up Rentrak, which before that bought Nielsen EDI to become the boss of box office tracking. According to comScore senior media analyst Paul Dergarabedian, who has covered the box office for 25 years (and missed just two Sunday reports in that time, he tells me), the vast majority of theaters’ data—at least 95 percent—is reported directly to comScore via point-of-sale systems. So when you buy a ticket at the counter or from a kiosk, the amount of your ticket purchase is sent directly to comScore. Some old-school theaters send their results to comScore by fax and a few mom-and-pop theaters prefer to call them in. Once in the system, the studios can log in to comScore’s database and watch in real time as the money pours in.

On Sunday mornings, weekend results are estimated based on the real numbers that are already in; actual weekend totals are ready on Monday. Hollywood sites like Deadline fight to get the numbers earlier than the competition, but they get the final numbers from the studios, with all the data coming through comScore first. “We are literally the guardians of this data,” says Dergarabedian. “You can go to all these sources like Mojo and all these other places, but the fact of the matter is we’re the only one that collects it for the studio to even report that number.”

If the weekend totals are the equivalent of a basketball box score, what are advanced box office analytics? Inflation-adjusted grosses are akin to adjusting a team or player’s scoring to the pace of play. Points per game can be compared to an actor’s gross per movie. In basketball, it’s not just the greatest player that matters, but the greatest dribbler, the greatest 3-point shooter, the best block percentage, true shooting percentage, and rebounds per 36. With movies, you can narrow them down to the biggest action buddy comedy movie (Men in Black, followed by Rush Hour 2), biggest R-rated film (The Passion of the Christ, then Deadpool), and biggest Thursday opening (Revenge of the Sith, then The Matrix Reloaded).

Patrick Ewing or John Stockton or Charles Barkley may be the greatest player who never won a championship, and Sing is the biggest box office earner that never led a weekend. How about the biggest Friday-to-Saturday drop percentage? Johnny Knoxville’s The Ringer (no relation) at minus-72.31 percent. Want to get more obscure? Jerry Maguire holds the distinction of being the best of the worst of the best by topping a weekend with the lowest amount since 1997 ($5.5 million in its seventh weekend, January 24-26, 1997). The biggest domestic weekend for a movie based on a play is Madea Goes to Jail at $41 million.

Movies with large “multipliers” are seen as particularly impressive by box office nerds. Get Out, The Greatest Showman, Titanic, and My Big Fat Greek Wedding are well-known examples. Each of them opened relatively modestly, but then stayed strong week after week. Get Out made 5.3 times its opening weekend gross. The Greatest Showman opened at just $13.4 million, but has since earned a total of $174 million, for a 13 multiplier. Titanic opened at $28.6 million and then finished at $601 million for a near-21 multiplier. The biggest three-day weekend for My Big Fat Greek Wedding was its 20th at $11.1 million—it finished with a whopping $241.4 million.

Fast & Furious is a fascinating franchise, kind of like Gonzaga in that it started out as an unknown, busted out of the gates unexpectedly, and kept up the momentum until it became a powerhouse. The first film made $207.3 million at the worldwide box office, about the same as the modest hit Horrible Bosses, and is currently ranked 715 on the all-time chart. The second film made a bit more. Then the franchise had an off year with Tokyo Drift. But every movie after that, through the seventh, made more than the last. The last two installments are in the all-time top 15 worldwide, making more than a billion dollars each. The same way Gonzaga became UNC, the Furious franchise is basically Star Wars now.

Movies, like teams, have preseason expectations. A Star Wars film is supposed to make bank, and when it doesn’t—like this year’s disappointing Solo—it’s considered a huge disappointment. But when a little movie with low or no expectations, like Get Out, breaks onto the national scene, it’s a feel-good story. (Solo has made a lot more at the worldwide box office than Get Out, but far less than any other Star Wars movie.) And just because a movie opens no. 1, or even tops a year, doesn’t mean it will win the battle for historical supremacy. Spider-Man 3 was the biggest domestic earner of 2007, but the comedies Superbad, Knocked Up, Ratatouille, and Juno are far more admired today.

Sometimes the box office, like the NBA, is about opportunity. Steph Curry, a relatively short and slight NBA athlete, may not have been a superstar in the days before the 3-point line. Likewise, Avatar may not have been the biggest film ever (or even among them) had another film beat it to the punch on 3D technology.

As with traditional sports coverage, box office prediction power has become a cottage industry unto itself. Shawn Robbins, chief analyst at BoxOffice Pro (and founder of The Box Office Theory, a forum dedicated to “statistics-driven box office commentary”), says prediction has become more of a science in recent years. A decade ago, predicting movie grosses “was more of a guess,” he says. But that has changed. “It’s far more data intensive,” says Robbins, “because there’s so much more information out there.”

Social media provides metrics based on the number of conversations about a movie happening on Twitter and Facebook and the amount that megastars with tens of millions of followers are promoting their films. Meanwhile, prerelease market research measures the public’s awareness (total and unaided) that a film is coming out, “definite interest” in seeing the film, and whether that movie is a surveyed person’s first choice of the films available. And all of it is broken down by demographics. Before It was released last fall, Deadline reported that “males overall are 24% unaided, 75[%] total awareness, 57% definite and 17% first choice. Females are 26% unaided, 76% total, 50% definite and 18% first choice. Teen males are 24% unaided, 75% total, 66% definite, and 24% first choice.” With not a single typically bankable star in the cast, It became the biggest unadjusted horror hit of all time.

Robbins likens the new data to that available in baseball, whose rise as a statistics-driven sport was made famous by the Oakland Athletics’ 2002 Moneyball season. “Box office has kind of gone through a similar phase,” says Robbins. “Certainly not as highly publicized and not as data-rich as a major sport is, but relatively speaking — whereas we just had more comparisons and intuition 15 years ago.”

Fandango provides a weekly “Fanticipation” report that is a useful indicator of what films will do at the box office. While it doesn’t include actual presale numbers, the report provides context for how well a major release is selling before opening. A few hours before Black Panther hit theaters, Fandango announced that it had already outsold presale runs of Captain America: Civil War, Deadpool, and The Hunger Games, strongly hinting that it would open in all-time fashion. Black Panther received a 99 out of 100 Fanticipation score, which is “calculated via an algorithm of Fandango’s advance ticket sales, website and mobile traffic, and social media engagement.” Avengers: Infinity War received a 100 and opened with the biggest weekend of all time (unadjusted for inflation).

Aggregated review sites Rotten Tomatoes and Metacritic, as well as post-screening audience surveys from comScore, can provide further clues to a movie’s box office potential, especially of multiplier, which is often related directly to word of mouth. The Greatest Showman has a Rotten Tomatoes review score of only 56 percent, but its audience score is 87 percent. And then there is the anticipation we can see: YouTube shows the number of views of an upcoming film’s trailer, an obvious but helpful indicator.

Box Office Mojo is owned by IMDb (which is owned by Amazon) and so it draws on IMDb pageviews to find comps. Just before Incredibles 2 opened, Mojo’s Brad Brevet wrote that in the two weeks leading up to release, it was “vastly outperforming” fellow animated sequels Finding Dory, Minions, and Monsters University in pageviews. It went on to open far bigger than any of those smash hits.

Predictors also factor in major weather and sporting events, and they look to school calendars to see how many college students and children will be in class or free to go to the movies. To predict international openings, they’ll look at the number of markets in which films are opening. (It’s rare to see a simultaneous “day and date” opening in every country.) The Chinese market is still protected by local regulators, so it’s not necessarily audience demand that dictates which American films open there at all, nor for how long they stay in Chinese theaters. The Marvel Cinematic Universe superteam, for example, does relatively better abroad than the Star Wars superteam.

Karie Bible, a box office analyst with Exhibitor Relations Co., which provides long-range forecasts and analysis to studio and theater-owner clients, says that for all the data available, the box office still turns expert predictions on their heads all the time. She points to Black Panther. “People thought it would do well, but no one thought it would be at almost $700 million domestic,” says Bible. “Sometimes things just happen that we don’t see coming in life and in the movies.”

Few could have predicted the box office take of Solo, which has done about one-third the total business of Rogue One: A Star Wars Story, its most comparable movie. And despite advances in box office analytics, there are still as many unpredictable factors in film as there are in NBA draft picks. Nobody guessed that Markelle Fultz would forget how to shoot or hurt his shoulder or whatever happened, or that Kevin Spacey would be replaced by Christopher Plummer after the trailer for All the Money in the World had already been released. It’s a safe bet that if Disney had known how popular Black Panther was going to be, its characters would have had even larger roles in Infinity War.

Though film companies are starting to hire analytics gurus from the sports world (notably Legendary), and using machine learning to predict a movie’s success from the screenplay alone, long-term box office predictions are still an inexact science. That’s because it’s not just a numbers game. “There is such a romanticism around movies,” Dergarabedian says. “It’s not like you’re selling widgets.” The movies, like sports, he says, are for the head and heart. To know what really happened in both, sometimes you have to watch the game.

That’s why someone with no connection to filmmaking can root for a movie, at its essence just another corporate product, to succeed. That’s why you can want Black Panther to “defeat” Infinity War and The Last Jedi at the box office, even though they are all owned by the same shareholders. And that’s why when The Crow beats 3 Ninjas Kick Back it feels like my favorite team has won the World Series.

Shaun Raviv is a freelance journalist based in Atlanta. This is his first story for The Ringer.

Shaun Raviv