TORONTO -- An analyst and author who accurately predicted 20 out of 21 Oscar predictions in 2018 is sharing one of the methods he uses to select the winners.
Ben Zauzmer, a Harvard graduate of applied math, found a way to combine his passions for math and films to create an Oscar predictions algorithm. In his book, 鈥淥scarMetrics: The Math Behind the Biggest Night in Hollywood,鈥� Zauzmer uses data collected from past awards shows and movie review sites such as Rotten Tomatoes to make his Oscar picks.
Zauzmer uses his formula to predict winners for 21 out of the 24 Oscar nominations; not including the three short-film categories because of the lack of statistical data on those films.
He says one way to make accurate predictions is by comparing films that are nominated in multiple similar categories. For example, frontrunners for best director are most likely to win best picture.
鈥淏est director has a really strong correlation with best picture and best screenplay, but the book found that best film editing is also a sneaky good predictor for best picture,鈥� Zauzmer said Thursday on CTV鈥檚 Your Morning.
As for this Sunday鈥檚 92nd Academy Awards, Zauzmer predicts the British war film, 鈥�1917鈥� and the dark-comedy 鈥淧arasite,鈥� are the two leading films to take home the major awards of the night.
The 鈥淥scarmetrics鈥� predict 鈥�1917鈥� with a 37 per cent chance to win best picture and its director Sam Mendes with a 54 per cent to take home best director. 鈥淧arasite鈥� is not far behind with a 19 per cent chance for the film, and a 20 per cent chance for its director, Bong Joon Ho.
Zauzmer says in the past, his favourites to win have won 77 per cent of the time. In 2018, he predicted 20 out of 21 correct. The one that got away that year was his pick for best documentary feature, French film 鈥淔aces Places鈥� which lost to Netflix鈥檚 鈥淚carus.鈥�
While his strategy has been accurate on most occasions, Zauzmer says it cannot predict everything; particularly the year 鈥淢oonlight鈥� won the Oscar for best picture in 2017 over the forecasted winner, 鈥淟a La Land.鈥�
鈥淭his is one of the classic upsets in Oscar history and certainly of the last decade. 鈥楲a La Land鈥� had won the BAFTAS, the Producers Guild, the top awards from the Golden Globes and yet it did not win best picture. This is why math doesn鈥檛 predict zero per cent or 100 per cent for any nominee. Upsets do happen and it鈥檚 what makes the Oscars exciting,鈥� Zauzmer said.
