Statsgasm: 2017 Best Picture Voting Simulation

Fellow Oscar watchers and numbers obsessives, I am happy to bring back Awards Daily’s Statsgasm for the 2017—2018 Oscar season. This year has been the most unpredictable I’ve ever seen in recent memory, and unless the guilds flip the script and all rally behind one film over the next few weeks, I think the safe bet would be to expect this wild and chaotic ride to continue all the way up to Oscar night, March 4th.

With nominations for the 90th Academy Awards just a few days away, Statsgasm returns with its fourth attempt at modeling the Best Picture nominating process, which as we all know centers on a quirky beast called the preferential ballot. A brief review: the Academy switched to a preferential voting system for determining its Best Picture lineup in 2009 — a move necessitated by the doubling of the number of Best Picture nominees from five to 10 films. Voters would be able to rank up to 10 films in order of preference. In 2011, the Academy modified this system so that voters would only rank up to five films and that there would be a variable number of nominees (between five and 10), depending on how ballots were distributed after three rounds of voting.

Although the Academy has been using the current voting system for six years, I’ve found that very few Oscar watchers have demonstrated a good grasp of preferential voting. Even fewer have dared attempt to model the process, which is a shame because to truly understand something like the preferential ballot, I feel that “learning by doing” and graphical demonstration is really the best way to proceed. As such, I do take some pride that Awards Daily is home to two different, detailed simulations of preferential balloting: Statsgasm (which uses year-end Top 10s from critics/bloggers as proxy ballots) and Dr. Rob Y’s Simulated Oscar Ballot (which asks readers to submit their choices).

As before, this voting simulation is *not* at all intended to predict the Best Picture nominees or how Academy voters will behave — it goes without saying that critics and bloggers are not industry voters. Rather, Statsgasm is only meant to give some insight into the pistons and gears driving the preferential system and how certain outcomes may occur.

Without further ado, here is Statsgasm’s Best Picture voting simulation for 2017, again powered by Tableau! (NOTE: for the best viewing experience, please use a tablet or computer to view the dashboard).

For the first time in four years, Statsgasm has arrived at what some have deemed impossible with the current system — a full 10 film lineup for Best Picture! The nominees are:

• Baby Driver
• Blade Runner 2049
• Call Me By Your Name
• Dunkirk
• The Florida Project
• Get Out
• Lady Bird
• Phantom Thread
• The Shape of Water
• Three Billboards Outside Ebbing, Missouri

Some notes and observations:

• This year demonstrates what I term as the “Goldilocks” vote distribution needed in the first round so that a 10 nominee Best Picture lineup has a chance of happening under this system. #1 votes were not heavily concentrated with one or two films as in years past, and only Get Out had enough first round support to trigger the surplus rule. In other words, those all-important #1 votes were more evenly divided among multiple films. Passion was spread out.
• Part 2 of the “Goldilocks” distribution: this year’s results also owe a lot to the fact that Call Me By Your Name was nominated at the end of Round 2 after getting just enough of Get Out’s surplus vote to hit the 9.1% (48 votes) victory threshold. If this hadn’t happened and Call Me By Your Name was still competing in Round 3, it would have taken a significant number of votes that other films needed to get nominated.
• The 3% rule I’ve observed since I began performing these simulations continues to be an excellent rule of thumb in determining the minimum amount of #1 votes a film needs to have a shot at getting nominated. This year, a record 10 films cleared 3% support (16 votes) in Round 1 and another was just a couple votes off. With one exception (explained by the controversial classification of Twin Peaks: The Return as a “film” by a very small group of voters who also ranked it first on their ballots), all of these films ended up being nominated by the end of Round 3. It was thrilling to count ballots from eliminated films in Round 3 and to root for Three Billboards and Baby Driver as they inched closer and closer to hitting the magic 5% number. In the end, both made it in by the skin of their teeth.

In all, I’m very excited by this year’s results, and although I don’t expect the Academy will follow suit with 10 Best Picture nominees next Tuesday, seeing it happen in a simulation gives me hope that it can happen eventually under the current implementation of preferential voting.

Please feel free to ask any questions in comments or directly! And if you haven’t done so already, here’s your reminder to enter Awards Daily’s “Predict the 90th Academy Awards Nominations” contest.

Happy predicting, friends, and stay tuned for more Statsgasm in the coming weeks!