Four Stars: Who Are Movie Reviews For?

Brandon Barrera
March 1, 2019

Watching a recommended movie is risky business. If the stars don’t align in your favor, you might find yourself nurturing a distrust of your source, forever altering conversations with friends and colleagues. Even when Oscar season rolls around, which should reliably provide lists of “good” movies, you might question if everyone sat through the same movie after scanning a few social media feeds. Does data science offer us evidence of something we might be missing?

PascalWallisch1-PhotoCredit(Yadin Goldman)
Pascal Wallisch, Ph.D.. Photo credit: Yadin Goldman

“There is a tremendous diversity in appraisal for any given movie,” said Pascal Wallisch, Ph.D., clinical assistant professor at NYU. “It’s actually quite striking.” Wallisch, seeking to measure the reliability of movie critics, gathered ratings from critics, aggregator sites (think Rotten Tomatoes and The Internet Movie Database (IMDB)) and a multi-year study with 3,000-participants. After determining the correlations of reviews from a pool of over 200 movies, he admits to being astonished—there was not a single film with any hint of a “moderate degree of agreement.”

“The Science of Movies,” presented by Wallisch and organized by Think&Drink NYC’s Gil Avidor, is a stimulating yet relaxed evening talk, suitably tailored to seekers of intelligent nightlife. Wallisch, whose research interests hone in on the intersection of psychology and neuroscience, extolled the virtues of finding your “movie twin,” bemoaned the scarcity of originality (ahem, creativity) in present-day Hollywood, and explained what happens to a brain exposed to a healthy dose of M. Night Shyamalan.

So what is a neuroscientist to do when exploring the disparity between average cinema-goers and taste-making critics? You put them in a scanner and have them watch 2004’s mystery-thriller The Village.

For the experiment, Wallisch’s team selected the relatively obscure and mixed-reviewed M. Night Shyamalan film to study its effect on the brain. The participants, having never seen the movie before, showed synchronicity in brain activity at the back of the head, where simple processing occurs. Conversely, the frontal cortex, specifically regions associated with the intrinsic system and default-mode network, appear to be areas where disengagement is detectable. “Disengagement predicts on a participant-by-participant level,” Wallisch said, “how much they like the movie, or not.” According to Wallisch, the scanner “sees” when participants stopped believing the premise and began zoning out.

While movie appraisal appears to reflect individual make-up, it is highly consistent within the individual. As practical advice for moviegoers, Wallisch suggests finding and following someone who shares their taste to a high degree, in essence their “movie-twin.” This could seem a little counter-intuitive to film buffs that rely on review aggregators like Rotten Tomatoes and are familiar with the conventional bifurcation between critics and average joes, the latter of which they might shrug off.

“Movie critics can predict other movie critics’ [ratings] pretty well,” Wallisch said. Not so much for the non-critic who might be contextualizing things differently. The correlation between a critic’s rating and a non-critic’s positive reception of a movie is quite low. It is statistically indistinguishable enough that a random person on the street could be as predictive as “Roger Ebert himself.” In other words, if you’re not a movie critic and “going in blind,” you are best served by trusting a film’s IMDB rating as a raw indicator of how much you can expect to enjoy a movie. Go figure.

Until a savvy entrepreneur makes FilmTwin a reality, Wallisch mulls over the question of whether creativity is still embraced in Hollywood. For anyone experiencing “sequel fatigue,” the seeming absence of original ideas and storylines in Tinseltown might come as no surprise. Of the movies at the top of the box office between 1975 and 2015, a smaller and smaller number of original intellectual properties have crossed the threshold. From a high mark in 1975 of all top 20 box office movies not being sequels, franchises or remakes, the percentage of original movies in the top box office of 2015 dropped to only 35%.

line chart, line sloping down over time
Figure 1. Pecentage of ‘original’ movies between 1975-2015. From: @Pascallisch

Commenting on the sustainability of the situation, Wallisch notes that television, with its long-form narrative, appears to be where creativity has fled. ”Can we make Hollywood great again?” Wallisch muses.

Brain Awareness Week is only a few weeks away! Visit our Calendar of Events to search for brainy activities in your area! For more NYC intellectual happy hours, browse through the Think&Drink NYC events page to buy tickets and visit braiNY for a Gotham-centric list.