Sunday, February 12, 2017

The VFX Predictinator, 89th Academy Awards Edition

What is The VFX Predictinator? Start here.

After bloviating in two giant posts about the reasons we might have incorrectly predicted the Visual Effects Oscar winner last year, we have decided to run the numbers for the 89th Academy Awards, perhaps for the last time:

  • 5.48 points for “The Jungle Book”
  • 4.19 points for “Rogue One: A Star Wars Story”
  • 3.87 points for “Kubo and the Two Strings”
  • 3.17 points for “Deepwater Horizon”
  • 3.09 points for “Doctor Strange”

The charred, withered, barely-breathing remains of The VFX Predictinator predicts that “The Jungle Book” will with the visual effects Oscar in the 89th Academy Awards. Here’s the full set of scores:

Directed by Jon Favreau, “Jungle Book” is a photorealistic live-action film that is nearly fully computer generated. The live-action portions of the film featuring young actor Neel Sethi were filmed on soundstages in Los Angeles. Computer graphics dominate each “Jungle Book” frame in, arguably, the most photoreal, most extensive, and most immersive use of CG environments and characters in a single film. “Jungle Book” advances upon the success of Oscar-winning “Life of Pi” (2012) to a staggering degree. Oh, and all the animals in the film talk, too. (Some even sing.)

“Jungle Book” took the most points in a competitive field this year; only 1.29 points separate “Jungle Book” from its closest competitor, “Rogue One”. Let’s take a look at the Predictinator criteria and how the race worked out for the five films.

On a number of fronts, the five Oscar nominated visual effects films are quite competitive. Most significantly, the five films all scored similar Tomatometer ratings from critics. In fact, these five films were the highest rated movies for critical acclaim for the 27 years of Oscar races that we’ve tracked. “Kubo” took home the most points for Critical Acclaim (with its 97% Tomatometer rating), while the lowest rated film was “Deepwater Horizon” with a very strong 83%. While this criteria was competitive, the scores for this category are relative to one another, so “Jungle Book” earned .75 points more than “Deepwater”.

None of the five films took away more than two Oscar nominations. As a result, no film earned any “Academy” points (a film starts earning “Academy” points only if it earns at least four Oscar nominations). Also, none of the five films were penalized for being a sequel; sequels, historically, are shunned at the Academy Awards. As you may have noticed, we did not label “Rogue One” a sequel. Even though the film is part of a cinematic universe, we decided that since the film follows a new set of characters, rather than a returning set of characters going on another adventure, it does not earn the sequel identity.

As an aside, if we continue running the Predictinator past this year, we are considering clarifying this piece of criteria to possibly include “reboots” and “films in an established cinematic universe”, since we feel like those films are also historically shunned at the Academy Awards.

With $364M domestic box office take, “Jungle Book” was the #2 earner among nominees (after “Rogue One”’s $522M), which decimated the box office point values for “Deepwater” and “Kubo”, since we score box office points relative to one another.

“Jungle Book” finished with a score 1.29 points higher than its next closest competitor, “Rogue One”. Most significantly, “Jungle Book” earned an important point for its primary visual effects consisting of organic creatures, and an additional .75 points for facial acting. None of the other four visual effects nominees earned any points in either of these two categories. The magnificent organic character animation in the film gives the movie a leg up, since, historically, the full Academy favors films with synthetic character animation and, particularly, animation that features characters that talk and emote.

The digital human work featured in “Rogue One” we deemed as supporting visual effects, not the primary visual effects created for the film. Space battles, environments, and “the world of Star Wars” are the primary vfx work of the film. We gave the same ruling last year to “The Force Awakens”, which featured computer generated characters Maz and Snoke as supporting visual effects elements.

Similarly, we also deemed the visual effects of “Kubo” to not qualify for ‘organic creature work’ as its primary visual effects. “Kubo”’s nomination for visual effects, the first for an animated film since 1993’s “The Nightmare Before Christmas”, is not without controversy. While live-action and animated films share techniques and concepts, their visually-oriented goals are substantially different. The vast majority of live-action films contain visual effects whose goal is to trick audiences into believing the sequences actually happened in front of real-life cameras, next to real human beings. Animated films do not share the same goal of tricking the audience in this way. Nevertheless, “Kubo” is a nominee, and just as we ruled for “Nightmare”, we deemed the primary visual effects of “Kubo” not to be the character animation in the film, but the creation of the entire world of the film. The art direction, creation, animation and assembly of the entire frame is the visual effect.

You may have noticed that we haven’t yet discussed “Doctor Strange”. Two years ago, we added the ‘is the film based on a comic book?’ criteria, which deducts one Predictinator point. “Strange”’s point value sits it solidly in the middle of the pack, just as most comic book movies typically do. To reiterate our reason for adding this criteria: comic book films have, historically, not been rewarded with visual effects Oscars.

We recently went into great detail discussing how our formula may be outdated since we may have already entered a new era of visual effects that renders our old assumptions no longer valid. We also qualified our argument with the “fluke year” defense: “Sometimes the 6000+ members of the Academy think differently than is expected; typically, the very next year, they go back to voting the way that is typically predicted.” This is the main reason we decided to run the numbers for this year’s race; if last year was, indeed, a fluke, then our philosophy remains sound.

Generally speaking, we feel pretty good about “Jungle Book” winning the Oscar; it’s the film that our gut tells us will win. It’s a nearly-universally loved film, with groundbreaking, well-executed visual effects. It was a giant hit, and Academy voters will feel good about rewarding this type of film.

That said, my wife and I have nagging concerns. Putting it mildly, last year’s “Ex Machina” win, which destroyed our formula, burned us pretty badly. As a result, we can’t help but look for potential spoilers, and attempt to pre-explain their potential victories.

The two potential spoilers of this race are, in our mind, “Deepwater” and “Kubo”. The two films are the least “Hollywood” of the five films, and considering the “Ex Machina” win from last year, we need to pay attention to these two well-regarded films.

“Deepwater” features a man vs. nature narrative (as opposed to the typical protagonist/antagonist structure of the three other nominees), whose visual effects strongly support the narrative, rather than serve the purpose of pure spectacle. Academy voters sometimes shun typically structured films with “good guys/bad guys”, especially when there’s a worthwhile alternative worth rewarding (“Hugo”, “Benjamin Button”, “Gravity” and “Life of Pi” are recent examples that come to mind).

 “Kubo” is a strong contender as well, considering the film represents a magnificent achievement in hand-created stop motion animation (along with a healthy, significant amount of computer graphics and digital compositing). Academy voters could vote for “Kubo” as a protest against computer graphics; “CGI is ruining movies!”, a misguided trope, still has traction in 2017. Even though a predominantly stop-motion animated film like “Kubo” couldn’t exist without modern technology, average Academy voters could reward the film as the anti-CGI nominee.

And, had “Arrival” made it past the bake-off to become a nominee, I’d be shouting from the rooftops that it could easily become this year’s “Ex Machina”. If “Arrival” earned a Visual Effects Oscar nomination, it would have earned nine Oscar nominations, and been a force to be reckoned with. (No, we are NOT going to run the numbers with “Arrival” as a nominee. It’s a lot of work. Feel free to do so on your own.)

For only the second time in the last nine years, none of this year’s visual effects nominees also have a Best Picture nomination. In contrast, last year, three out of the five nominees were nominated for Best Picture (“Revenant”, “Mad Max: Fury Road” and “The Martian”). In addition, only one out of the ten films that participated in the visual effects bake-off was ultimately nominated for Best Picture (“Arrival”).

Marvel Studios earned its seventh visual effects Oscar nomination in nine years, with “Doctor Strange”. As a reminder, Marvel Studios debuted nine years ago with “Iron Man” (2008).

We’ll see what happens when the 89th Academy Awards take place on February 26, 2017.

Update: We got it right.

1 comment:

darkmoon3d said...

I'm feeling same with 'Jungle Book' winning. All around favorite pick from everyone. Wondering if last year's fluke year might've been because of most nominations tying so close with high predicinator points. Voters probably voted for the odd one out. Has there been any other really close predicinator scores? This is the only legitimate explanation I can think to fix predicinator for that year.