Box Office Trackers Can’t Crack the MCU Code

In its second weekend in theaters, Avengers: Endgame became the fifth film ever to top $2 billion in worldwide box office earnings, edging out Titanic for the second spot on the non-inflation-adjusted leaderboard. Endgame needed only 11 days to reach the $2 billion mark, making it the fastest film to do so. Only Avatar at $2.8 billion still stands ahead, and Endgame, which trailed by $600 million on Monday, has much better than a 1-in-14-million chance to surpass it, despite its even more protracted running time and less ideal launch date.

No one is surprised to see Endgame minting massive amounts of money, but the extent of its earning power is stunning nonetheless. Endgame’s opening weekend gross—which incredibly accounted for nearly 58 percent of the industry’s domestic intake across all movies—leveraged a simultaneous launch in China to nearly double the previous largest worldwide gross, set by its 2018 predecessor, Avengers: Infinity War. Endgame’s earnings also outstripped the previous record domestic take (again, Infinity War) by almost $100 million, lapping the field even in inflation-adjusted dollars.

Projections published in the days and weeks leading up to Endgame’s release made it clear that the industry had more modest expectations, if you can call numbers that would rival the biggest openings ever modest (and in this case, you can). On April 4, a few weeks before Endgame opened, Variety reported that the movie was “on pace for a debut between $200 million and $260 million, according to early tracking.” On April 23, a few days before the April 26 public premiere, that estimate had climbed and narrowed to “between $250 million and $268 million.” Yet the movie’s actual domestic total came to more than $357 million, roughly 38 percent more than the midpoint of that pre-release estimate.

Although Endgame overperformed to an extreme degree, blowing by projected opening earnings has become commonplace for Marvel movies. Captain Marvel’s early estimate pegged it for a $100 million opening, but the movie ended up making more than $153 million. Although Ant-Man and the Wasp’s $75 million came in right around publicly reported projections, Avengers: Infinity War earned almost $258 million after estimates in Deadline and The Hollywood Reporter timidly suggested it could reach “$200 million-plus.” THR reported that tracking service NRG had brightened its forecast for Black Panther considerably over the few weeks before the film’s release date, but the projected figure as of a few days before the opening still undershot the eventual earnings by about 22 percent. Thor: Ragnarok was reportedly tracking at $100 million to $110 million but finished at almost $123 million, and Spider-Man: Homecoming was expected to earn $85 million to $110 million but wound up at $117 million.

The most recent movie in the Marvel Cinematic Universe to underperform reported projections was 2017’s Guardians of the Galaxy Vol. 2. That was eight MCU releases ago, and it missed by only a few million. Most recent entries in the most remunerative movie franchise of all time have been so successful financially that the industry has struggled to recalibrate the baseline for what Marvel movies make. Even in an era when watching the box office has become a sort of sport and companies are increasingly applying algorithmic methods in order to forecast films’ earnings, the MCU is a juggernaut that defies prediction. No matter how high the projections go, the actual earnings tend to go higher.

NRG declined to provide its past projections for MCU films, but two other providers were willing to share. Ben Carlson, the cofounder and copresident of Fizziology, is one of the professionals charged with the difficult task of projecting how much Marvel movies will make. His company provides box office estimates and tracking for the market research firm MarketCast, which acquired it in 2017. Fizziology’s numbers are predicated entirely on social media data—drawing largely on Twitter, Facebook, and Instagram, among other sources—and incorporate “volume, sentiment, geography, gender, and intent” as predictive factors.

Carlson confirms that the rarefied financial air in which the MCU operates poses a unique predictive challenge. It’s easier to anticipate a movie’s earning power, he says via email, “when a new data point is surrounded by dozens of others on all sides. You can then say things like ‘90% of films that look like this open between $30-$40m domestically.’ But these films are so much bigger—by every metric—that there aren’t any data point comps to point the way. So they end up being a [shrug emoji].” With great projections comes great uncertainty; the smaller the sample of previous movies with similar online buzz, the bigger the error bars surrounding the projection. And very few movies are as buzzed about as Marvel’s.

Carlson sent Fizziology’s prerelease projections for the past five MCU movies, displayed in the table below. Although the figures are closer than some of those reported in the prominent trade publications, they’re still consistently low, with the exception of Ant-Man and the Wasp, which was “normal” enough to fit in with a non-MCU crowd. The plus signs on the other projections, Carlson says, indicate that “we knew the floor, but didn’t know how high the ceiling would be.”

“The biggest Marvel movies—especially the Avengers films—were so big that we had no comp data to guide us,” Carlson says, “leaving us knowing that they would be huge, but unsure of just how big they would be.” With the high side obscured, the projections appear to play it safe and look low in retrospect.

The MCU box office effect holds up even according to a different projection method. Joshua Lynn is the president of Piedmont Media Research, a company that predicts box office performance using a neural network that includes a consumer engagement metric—which gauges the audience’s prerelease response to the movie’s concept and cast and, Lynn says, has a 0.74 correlation to opening-weekend box office—as well as more concrete factors such as production budget, screen count, Rotten Tomatoes score, and audience reception to early screenings.

Piedmont’s database contains all but the first five MCU films, beginning with 2012’s Marvel’s The Avengers. On average, the MCU movies have outperformed Piedmont’s projections by about $18.3 million, or 13.6 percent. This trend doesn’t extend to all superhero films, or even all films featuring Marvel characters. The 10 non-MCU Marvel films in the database, which range from 2012’s The Amazing Spider-Man to 2018’s Venom, actually underperformed, on average, by a bit more than $2 million, or 2.6 percent. The non-MCU films, which weren’t produced by Marvel Studios, averaged 23 points lower and 10 points lower on Rotten Tomatoes’ critic and audience scores, respectively, according to Lynn’s records, which explains some of Marvel’s stronger-than-expected grosses on its MCU releases. “By and large, they just make really good movies,” Lynn says.

Like Carlson, Lynn mentions the difficulty of projecting movies with few comparable productions, and he notes that we’re dealing with fairly small samples when we restrict our scope to Marvel movies alone. He also cautions against falling prey to recency bias. Only 10 of the 17 MCU films in the database exceeded expectations, but those that did beat their projections by an average of $39.6 million, or almost 30 percent, while the seven that underperformed fell short by only $12.2 million, or about 10 percent. (Eight of the 10 non-MCU films underperformed, with Deadpool the lone big breakout.) Three of the four biggest MCU overperformers—Black Panther, Infinity War, and Endgame—are among the most recent five films in the franchise.

The estimates Piedmont provides can fluctuate significantly as a movie approaches release and the audience assesses its quality. The consumer engagement score for Guardians of the Galaxy, Lynn says, spiked by about 75 percent when Piedmont reran it after the trailer appeared, because the characters weren’t well-known initially. And the much-maligned Fantastic Four, he adds, would have projected for a $70 million to $75 million opening (rather than a $28 million opening) had its Rotten Tomatoes score hovered in the 80 to 90 percent range (as do most MCU movies’) rather than the 9 percent it actually earned. If Marvel Studios were to produce a dud, that could end the pattern of overperformance, but thus far, no MCU movie has dipped below a 66 percent critic score (Thor: The Dark World) or a 70 percent audience score (The Incredible Hulk), save for Captain Marvel’s troll-tanked audience score. Most of the recent releases have comfortably beaten both of those figures.

Of course, Avengers movies have been the biggest MCU standouts, and those movies may not be a renewable resource. Lynn notes that another notable projection-topper, Black Panther, was part of “a larger cultural moment with certain groups that were really pushing and for the first time actually seeing themselves represented up on the screen in big ways.” That effect isn’t easy to model, and it may not be easy for future films to replicate. Endgame also opened on 188 more domestic screens than any earlier Marvel movie in Piedmont’s database, which probably pushed theaters to capacity and maxed out its earning power.

Although Endgame ostensibly brought an end to the Avengers saga, the MCU assembly line hasn’t slowed. On Monday, the internet quaked as it greeted the trailer for the next MCU movie (and the final film in “Phase Three”), Spider-Man: Far From Home, which will arrive in early July. Although we can’t yet say how the film (and its apparent reliance on the multiverse) will be received, Lynn generated an early projection based on Piedmont’s consumer engagement score for Far From Home, coupled with Homecoming’s budget, screen count, and Rotten Tomatoes scores. That produced a projection of $122 million to $133 million, slightly higher than Homecoming’s opening gross, which Lynn attributes to Far From Home’s “incredibly high scoring with key groups” (in Homecoming’s case, men 18 to 49 and teenagers of all genders). After 22 MCU entries with nary a bomb in the bunch, fans’ fondness for past Marvel movies—coupled with the continuity that ties each sequel to its predecessors and successors—makes the whole franchise function like a movie version of a perpetual motion machine. And every time an MCU entry makes more money than expected, its success spawns headlines that only add to the hype.

Given that one of the problems with projecting MCU movies is the paucity of comparable blockbusters, Endgame’s earnings alone may help balance the scales by raising expectations and making it tougher for future MCU titles to exceed them. “The more data that exists in the upper stratosphere of [the] box office, the better our ability to forecast massive openings,” Carlson says. “Now that the bar has been reset by Marvel—again—we should be able to better understand future performance. ‘Well, it’s not going to do Endgame numbers’ is something we can now say.” These days, though, forecasting an in-house Marvel movie not to make a killing would require some courage. Particularly in the past two years, taking the under on the MCU has been a very bad bet.