Monday night, the man who will, in less than two months, become the president of the United States, used his Twitter feed to disseminate baseless claims of voter fraud originating from a handful of obscure accounts, one of which belonged to a 16-year-old Anaheim Ducks fan. As he did during the campaign, when manual retweets of conspiracy theories from neo-Nazi accounts were a common occurrence, Trump showed no apparent concern for the credibility of his sources, but with @HighOnHillcrest, he inadvertently struck gold. The HighOnHillcrest tweet Trump chose to spread to his more than 16 million followers attacked CNN’s Jeff Zeleny for failing to prove that Trump did not “suffer from millions of FRAUD votes,” one of more than three dozen in a row attacking the news network, along with a wide assortment of journalists and publications. But scroll back a few days, and you’ll find HighOnHillcrest going after the real big game: Rotten Tomatoes. “Pls fix @RottenTomatoes Tomatometer,” he tweeted at the site’s editor-in-chief, Matt Atchity. “Percentages are misleading & don’t give REAL review summary. ADD rating percentage avg!” (Naturally, he tagged CNN as well.)
HighOnHillcrest may be wrong about the likelihood that millions of fraudulent votes were cast in the last election, and he’s certainly confused about where the burden of proof lies. But where movie review aggregation sites are concerned, he has a point. Rotten Tomatoes’ iconic Tomatometer is misleading, or at least it doesn’t convey what most people think it does, and the site does little to combat that misperception. We naturally interpret percentages the way we learned in grade school: Something in the mid-80s is a solid B, and 100 percent is absolutely flawless. But Rotten Tomatoes’ rating system doesn’t distinguish between a rave review and a lukewarm one. No matter how weak a critic’s thumbs-up is, it counts as “Fresh.” It’s the review-aggregation equivalent of the Electoral College, where winning by one vote or by millions produces the same end result.
As with the Electoral College, the rules by which the Tomatometer operates are clearly defined, so there’s no real excuse for misinterpreting the results, whether it’s thinking that a “100% Fresh” rating means critics passionately love a movie rather than that none of them dislikes it, or claiming that the third-narrowest Electoral College win since 1980 and a popular vote loss of more than 2 million constitutes a “landslide.” The problem is that what the Tomatometer actually reflects is less useful than what many people think it reflects. If you’re trying to choose what to see, you don’t really want to know how many critics think a movie is basically OK. You want to know if they love it and how much, in which case Metacritic, which draws from a more selective pool of critics and scores their reviews for content, not a simple thumbs-up or -down, is a far more effective tool. (Caveat: Both sites reflect the institutional biases of contemporary criticism, which is to say they’re heavily tilted toward white males and underrepresent women and people of color, and it’s always better to rely on the voices of individual critics than a numerical aggregate.)
On Metacritic, a score of 100 indicates not only unanimous critical approval but uniform adulation, a feat so rare that only three movies in the site’s history have achieved it. Rotten Tomatoes, by contrast, lists four movies with “100% Fresh” scores in 2016 alone. But Rotten Tomatoes has so dominated the discourse of review aggregation that you often see Metacritic’s ratings followed by a percent sign, even though that’s not what Metascores represent. Studios, naturally, prefer the more lenient Tomatometer—who wouldn’t opt to bill, say, The Jungle Book, as a movie that’s 94-percent fresh rather than one with “generally favorable reviews” and a Metascore of 77—and their use of the Rotten Tomatoes logo in ads and on DVD boxes further cements the site’s prominence. In other words, at least as far as the practice of aggregating reviews and converting them into a numerical value, HighOnHillcrest has a solid point. Now, about that voter-fraud thing …