Predictions for Oscar 2020: Joaquin Phoenix is ​​the sure victory, while "Parasite" is the dark horse

Segway-like vehicles were also used by the police to keep an eye on the large crowd on Marina Beach.

The Oscars are predictable. And there is a science behind such predictions.

While the level of accuracy varies from one prediction model to another, a candidate stands out in each category every year. For example, it can be predicted with 99.8% certainty that Joaquin Phoenix will be recognized as the best actor in a leading role this year joker, While in the category “best picture” the probability of winning is relatively closer to 71% at 1917, joker at 10% and parasite only at 2%.

But how do statisticians predict the Oscar winners? More importantly, how many past winners match the forecasts?

While the mathematical model contains complicated equations, the basic requirement is simple.

Have you ever played the battleship game? While the chances of spotting an enemy ship are small at first, the likelihood increases as the game progresses. The Oscar predictions follow the same principle: if an exercise repeats over a period of time, enough patterns are created to guess the future results. With the 92nd Oscar, awarded on February 9, mathematicians have 91 earlier results than data to create a model that can be used to predict the future winner.

Iain Pardoe, who holds a PhD in Statistics from the University of Minnesota, built such a model that has proven to be one of the most accurate.

He uses three basic patterns.

First, the members of the Academy who are involved in the decision about the Oscars are also represented on the jury that decides on the Golden Globes, the Directors Guild of America, the Screen Actors Guild, the BAFTA, etc. For example, 34 of the 64 winners of the best picture had previously won the Golden Globes from 1943 to 2006. The awards before the Oscars are therefore an important data point.

Second, Best Picture and Best Director winners are often represented by multiple nominees in other categories. For example, between 1938 and 2006, only three films won the best picture without being nominated for best director. The correlation between the nominations thus contributes to the ultimate probability.

Third, previous nominations of main actors and directors increase their chances of winning. While previous victories reduce the likelihood of repetition for starring and starring characters.

The weights are then assigned based on the assumptions above, which correlate strongly with previous Oscar winners. Using this relationship, the winners of 2019 are forecast. The table below shows the predicted winners in the four main categories – Best Picture, Best Actor in a Leading Role, Best Actress in a Leading Role, Best Director – and their odds of winning.

The winners are forecast using our data-driven model

The winners are forecast using our data-driven model

A word of caution. No model can be 100% accurate. Between 1938 and 2018, the selected model accurately predicted 71% of the Oscar winners.

With a success rate of 80%, the chances improve if only the last 14 years are taken into account.

However, the academy has provided surprises on many occasions. Especially the category "best picture" – since it represents a collective effort and not the brilliance of an individual – has often disappeared from the charts.

Notable exceptions in recent years: 2016 winners moonlight (only had a 2% chance of winning, but struck La La Land who had a 97% chance); 2015 Winner headlights (with only 6% probability hit The revenant who had 56% chance).

There may be exceptions to Korean film this time too parasite ride the race as a dark horse.