Data science and data analysis have been working hard in the business world for some time—something about which Grow is doing its part. More recently, though, sports teams have jumped on the analytics train to the extent that many—especially baseball—now have analytics teams in the dugout.
Fantasy footballers have also taken to using high-level analytics created by individuals and companies to determine how best to compose their fantasy team. (Click here to see Grow’s contribution to this.)
For the last few years—more for fun than anything else—some analytics and data science practitioners have adapted their skills to predicting what films will win at “the Oscars,” aka, the Academy Awards.
Back in 2016, according to a BBC business story, two technology companies claimed to know which film would win the Best Picture Oscar that year. The two companies, Cognizant and Clarabridge, predicted that The Revenant, with Leonardo DiCaprio, would take the Oscar.
They “reached their conclusions not by watching the films and applying artistic criticism, but simply by crunching data—lots of data,” according to the BBC article. “They looked at 150 variables, from film genre to box office takings, from review ratings to the percentage of female viewers under 18. And they applied their algorithm to data going back 15 years to work out which of these variables were the most important.”
They also looked at IMBD and Rotten Tomatoes to determine sentiment. They looked at 150,000 text reviews and more than 38 million star ratings from IMDB alone.
So after all this, the tech firms said they had 64 percent confidence that The Revenant would win. The article points out that, while this might not seem like a particularly high score, “their next highest confidence score was 19.2 percent for Mad Max: Fury Road, so the scores are relative.
Here are the final confidence ratings for that year:
So what happened? Well, for those who don’t recall, the film in the 6th spot—Spotlight—took the award.
An article in February of 2019 on Time.com suggested the winner for last year. A Time data writer teamed up with a professor of statistics at Virginia Tech who programmed a computer to predict the Best Picture winner. Their pick: Roma.
So what was their process? “Our oracle is a fairly simple computer program we wrote that accepts 69 years of data on major film awards, ignores all but the most predictive variables, and returns a statistical model that can predict past winners with extraordinary accuracy,” they explained. . . . “The model is still bullish on Roma. . . as the leading contender for the top honors, assigning it a score of 45.5 out of 100, which is four times higher than the closest competition.”
In descending order, here was their prediction:
“Green Book” won.
More recently, according to an article on ew.com, Ben Zauzmer is playing the Oscar-prediction game. Zauzmer, who works for the Los Angeles Dodgers as a manager of baseball analytics, “has for the past eight years taken his knowledge of applied math to create a model that predicts 21 of the 24 Oscar categories. . . .”
Without getting too far into the weeds, Zauzmer explains that he looks at Oscar data from previous years, including which categories every film was nominated in, and how all of these movies did in each category at the Golden Globes and the BAFTAs and the many other award shows.
“I have the scores of each movie on Rotten Tomatoes and Metacritic,” he said. “Anything that I can put into a number, I save on my computer. Then I can look at past years' Oscars to see how well each of these different things predicted each category. So maybe in one category, it’s the guild awards that are the best predictor. And then in another category, it’s the BAFTAs that are the best predictor, and so on. That gives me a set of weights, how much weight to put on each of these predictors. The ones that have done a better job of predicting the Oscars get more weight, according to various statistical formulas. And then when all is said and done, I can take these weights and apply them to this year’s nominees. And that’s what gives me the probability that each of the 2019 contenders will come away with an Oscar in each category.”
So what’s Zauzmer’s prediction? Well, at the time the article was published, his gut feeling was that it’s a race between The Irishman and Once Upon a Time in Hollywood. We will see.
What about this year? Well, we thought we might let Grow’s super-cool dashboard take a starring role by taking a whack at predicting this year’s Best Picture.
Well, not so fast, according to Grow’s Data Key Grip, Juliet Fletcher. On paper, she explained, 1917 appears to be the likely winner because of the numerous awards it’s already won: Golden Globes, Producers Guild of America, and British Film Academy (BAFTA). “Oscarmetrics” Zauzmer doesn’t think it’s so simple. According to an article on the LAist, it's an interesting year for making predictions, he said. There are “four movies receiving 10 or more nominations each for the first time ever. That could make determining the winners that much more difficult.” Joker has 11 nominations, and 1917, the Irishman, and Once Upon A Time In Hollywood all have 10 each.
Fletcher reminds us, however, that the Academy is now using a preferential ballot system to determine Best Picture, which has a tendency to change the calculus somewhat. Under this format, in order to prevent a less popular movie from winning a fractured contest, the voters rank the nominees from most favorite to least favorite.
So, based on all the data and balloting intricacies and, well, some gut, what does Fletcher and her Grow Dashboard have to say? “I’m picking Parasite, she announced, “which recently--and historically--took home the 2020 Screen Actors Guild (SAG) award for Best Picture.” (SAG also uses a preferential ballot, similar to the Oscars).
Fletcher has other reasons, too, some based on data, some not. “There is no doubt that, with all the talk about the lack of diversity in this year’s nominations, an international film winning would be a huge statement.” What’s more, she said, Parasite was also one of the year's best-reviewed movies--99 percent on Rotten Tomatoes, which was the the highest number of any Best Picture nominee.
In the end, Fletcher and her Grow Dashboard data pretty much agree with Justin Chang writing in the LA Times: "I know, I know: (Parasite is) a foolhardy prediction buoyed by wishful thinking. With awards from numerous precursor groups including the Producers Guild of America (which, like the motion picture academy, uses a preferential ballot), 1917 is clearly the statistical favorite. But the stats have mattered less and less in recent years, and, this year, Parasite, a movie with no small amount of guild support itself, has become the darkest of horses, the wildest of cards. Far more than last year’s Roma, Bong Joon Ho’s movie seems poised to become the first non-English-language film ever to win best picture—and what’s more, it deserves to be."
Fletcher agrees and wants to thank all of her people for their support.
See you on the Red Carpet next Sunday.