Tuesday, July 04, 2006

"Nobody Knows Anything" Proven?

Everyone knows the famous opinion of screenwriter William Goldman that, in Hollywood, "Nobody knows anything." It's often quoted to show the utter unpredictibility of the movie business, and to tear down any self-styled expert who claims to have an insight into delivering films that the public wants to see. In the LA Times West magazine last Sunday, Caltech professor and former screenwriter Leonard Mlodinow uses the mathematics of randomness to go a step farther and show that there's "no way studio executives or producers, despite all their swagger, can have a better track record at choosing projects than an ape throwing darts at a dartboard..." ...except that I don't buy it, at least when it comes to animated features. Here's his basic premise: "In science, data are not accepted as meaningful if they're the result of chance alone. People in the film industry are diligent about gathering data, but are far less skilled at understanding what the numbers mean. The fact is, financial success or failure in Hollywood is determined less by anyone's skill to pick hits, or lack thereof, than by the random nature of the universe. The typical patterns of randomness—apparent hot or cold streaks, or the bunching of data into clusters—are routinely misinterpreted and, worse, acted upon as if a new trend had been discovered or a new epiphany achieved. And so, despite a growing body of evidence that box-office revenue follows the laws of chaotic systems, meaning that it is inherently unpredictable, the superstructure of Hollywood's culture—that pervasive worship of who's hot and the shunning of who's not—continues to rest on a foundation of misconception and mirage." He goes on to give examples of producers who seemed to have hot streaks of sustained success, but then shows they all had "regressions to the mean," meaning in the long run their success was no more likely than that gorilla choosing films by randomly tossing darts. You see, a given producer only appears to be consistently having success because we misunderstand, and fail to appreciate, that it's really a random process. And in live action, the analysis is pretty convincing. But what of animation? If this is true in our industry, then in the long run we'll be able to recognize that John Lasseter will have the same success rate as Don Bluth or Ralph Bakshi. That it was a random process that DreamWorks and Katzenberg started having huge success when they gave up on serious-minded animated epics, and started doing raucous comedies. That Doogle or Hoodwinked had the same random likelihood of being smashes as Ice Age 2 or Cars or Over the Hedge. And that, of the films coming later this year, Open Season (my pick for the most successful animated film of the second half of 2006) is no more likely than Everyone's Hero or Ant Bully or Happy Feet or The Barnyard to win over audiences. Another quote from a mathematical expert in the article: "But a careful study reveals that no strategy the studios devise is going to give them any kind of advantage at all. So any studio executive getting paid more than the salary of a comparable executive at your local dairy is getting paid too much." Well, I would posit that after 100 years of animation, we do have some idea about what kinds of characters and films are more likely to appeal to audiences. And that films made to the highest standards, that ooze quality and loving craftsmanship, are more likely to succeed. And that Iger didn't waste billions buying the expertise of Lasseter and his team. Time will tell, but when it comes to animation, I'm not betting on the gorilla.


Mark Mayerson said...

Pinocchio was a flop. The Iron Giant was a flop.

The Care Bears Movie was a hit. Hoodwinked was a hit.

Do I think that the Care Bears Movie is better than Pinocchio? Obviously not. Therefore, something in addition to the quality of the films is determining their success.

That was the point of the article. There are so many variables that no matter how safe you try and play things by picking the right story, talent, etc. you cannot predict the outcome.

I think that applies to animation as much as it applies to live action.

Kevin Koch said...

The issue isn't what was a flop and what wasn't. The issue is, can we make meaningful predictions about what films are more likely to succeed? We know that there is a certain randomness to the process. I'm arguing, though, that the process is not completely random, as the article states.

Also, as I wrote, quality is but one part of the equation. As I alluded to, DreamWorks made a ton of high-quality flops and near-misses, because there was little attention to other key elements. Even as I worked on those films, I frankly predicted that most of them would fail -- not because of quality issues, but because some combination of the premise/setting/characters/story lacked appeal.

Also, the article is talking about success with audiences -- i.e., absolute box office. In those terms, calling Hoodwinked a success means that you also have to label Spirit and Road To El Dorado and Atlantis and Treasure Planet successes, because audiences responded as well, or even better, to all those films. I'm well aware that a part of real success is profit, so the production costs of a film matter, But for the purposes of this article, and my critique, we're talking about success with a wide audience. $50 million domestic box office is NOT why anybody is making these films. (And, by the way, if you count in the cost of redos after the Weinsteins bought Hoodwinked, and the cost of redubbing the entire film with a big-name cast, and the costs of distribution and a major marketing campaign, I'm not sure you can really call the film much of a success).

If we cannot predict the outcome in animation, then the success rate of Pixar is a statistical improbability that beggars the imagination. The the statistics of the turn-around at DreamWorks since the company not only went to CG but also completely changed their model for what kinds of films would be made also strain the idea of randomness.

But hey, maybe you're right. Maybe we'll see The Barnyard and Ant Bully perform as well as Cars and Ice Age 2. But I'll go out on a limb and predict that they won't.

Unknown said...

Actually "nobody knows anything" works the other way too - especially with Pixar. Before Nemo's release I heard plenty of people predicting it as a flop - even many who worked on it were secretly calling it Fishtar. I doubt even in Pixar's wildest dreams they ever would've predicted that it would have done the BO it finally did. Actually the same can be said for the LionKing - a film that was started and stopped and then remade with a new set of directors. If I recall right there were stories that many animators wanted nothing to do with the film.

My feeling is that you can get close and produce a film that should do well, but with the changes in the audiences and what other studios release at the same time or near or which team is in which playoffs, etc, etc. it really is impossible to predict a sure thing. There might be surer bets than others and Pixar sure is the surest bet there is right now - accept maybe Shrek3, but it still can all change on a dime.
I'm glad I'm not having to put up my money on whether the next animated film will be a hit or not.

Kevin Koch said...

I agree that it's impossible to predict a sure thing. But I think if you're reasonably savvy you can predict which films are likely to do well, and which films have little chance. It is not an utterly random process. For example, I'll predict right now that Ratatouille will do much better than Foodfight!. ;)

Mark Mayerson said...

Kevin, you're stacking the deck if you're comparing Ratatouille and Foodfight unless you're willing to deal with return on investment. There's no question that Ratatouille will have a bigger budget, better marketing and a more experienced crew. It's highly likely that it will gross more than Foodfight. However, it's quite possible that Foodfight will provide its owners with a higher percentage of profit.

Anonymous said...

(And, by the way, if you count in the cost of redos after the Weinsteins bought Hoodwinked, and the cost of redubbing the entire film with a big-name cast, and the costs of distribution and a major marketing campaign, I'm not sure you can really call the film much of a success).

That's an interesting take on "Hoodwinked" that I had not heard before . Most articles I've read about Hoodwinked have trumpeted how little the film supposedly cost and how much it made at the box-office (relative to it's cost) as evidence of what a great financial success it was.

Do you have some firm numbers on what additional costs there were to the Weinstein Co. for the re-takes and the big name cast re-dubbing on "Hoodwinked" ?

Kevin Koch said...

No firm numbers, but I do have some inside info on all that had to be done to the film after the Weinsteins bought the film. And I know the typical costs for striking prints and distribution, as well as the typical costs for an extensive marketing campaign. From what I saw, that was at least a $25 million marketing campaign. Plus, as we all know, the studio only gets 50-55% of the box office gross, so take that $50+ million and cut it in half. Also, the foreign grosses were minimal. So, basically, the film was still comfortably in the red until DVD/video.

Mark, the article talked about predicting what movies would be hits, not return on investment. They're not the same thing. The article is about predicting what films will catch on with an audience in a big way. According to the theory that "it's all random," a more experienced crew or better marketing are just variables that don't matter. And in live action, we can point to many, many films with huge budgets, top talent, huge marketing ... that completely failed. And much cheaper films, with lesser marketing, that hit big (in absolute numbers). However, in animation, I think those variables do matter -- which means, it ain't random!

Also, as the Hoodwinked example above illustrates, comparing production budget to final gross won't come close to giving you an idea on return on investment. It's long been known that a single very expensive smash hit is far more profitable than a whole slew of low-budget baby hits.

Anonymous said...

Returning to the top comment:

"Pinocchio" was a flop only in terms of "Snow White"...and it's production costs. Disney's first feature grossed $8 million domestic in 1938, a HUGE amount for the time, probably second only to "Birth of a Nation" for which no box office records exist, since it was sold state by state in 1915.

"Pinoke" did respectably at the box office, but it was the most expensive film -- dollar for minute of film -- made to that time. It was the most expensive pre-war Disney feature; although it cost less than, say, "Gone With the Wind." (2.7 million vs. $4 million) But GWTW was 3 hours and forty five minutes long. "Pinocchio" was 84 minutes.

The reason that "Pinocchio" was considered a flop in its first release was because it didn't return its (stratospheric) cost. But it did relatively well at the box office.

Jenny Lerew said...

Didn't Pinocchio also suffer largely because of the sudden war in Europe--and the loss of the international box office?

Kevin Koch said...

I think that's right.

The other part about doing high-quality animated features is, even if they aren't a big hit in their initial release, they still have a chance to make serious money in the long run. I know that's not a factor in the discussion here, but I think it's relevant (given the extra long life span of well done animation compared to live action). Walt Disney had plenty of films that performed poorly in their initial release, but they virtually all went on to make money. On the other hand, has anyone bought a Care Bears video or DVD in the last 10 years?

Mark Mayerson said...

Kevin, WWII is exactly the kind of wild card occurance that only bolsters the argument in the original article. When it takes so many years for an animated feature to make it to the screen, there are more things that can go wrong to affect the box office than in live action, not less.

The world where Disney could theatrically re-release animated features no longer exists. I suspect that the long term income potential of Toy Story or The Incredibles is going to be far less than Snow White or Pinocchio. By that I mean that after 20 or 50 years, it will be clear that Toy Story made the bulk of its money earlier than Snow White did.

Furthermore, the Care Bears have been "updated" to be cgi and there are new Care Bears direct to video features and a new cgi Care Bears TV series. If people weren't buying them, they wouldn't be making them. I hate the Care Bears, but if you add up all the revenue they've generated, you might discover that they're more lucrative than many animated features (merchandising included).

Kevin Koch said...

Mark, the explicit point of the article was that choosing/predicting films that are likely to be successful is completely random. My argument is that in animation, it is far from completely random. But I'm not arguing that it's binary (i.e., completely random or completely predictible). I think it's some where in between. I think a savvy, sensitive producer CAN choose projects that are far more likely to succeed than that gorilla with a dartboard. And I think our industry's history shows that.

So of course there are scores of huge variables, of which wars and, say, natural disasters are just two. But even with that, it ain't completely random. Especially because animation, much more than live action, gets "second chances" to make money.

Regarding the lack of theatrical re-release causing film's income to be more "front loaded" than in the past, I disagree. With video and dvd and pay-per-view (and, soon, on-demand at a small cost for every movie ever made), films continue to make money loooong after their initial release. If you look at the actual theatrical grosses for most of those Disney re-releases, it wasn't that impressive (and of course they had to share 50% of that income, plus pay new marketing costs, etc.). Yes, it added up in the long run, but we're talking incremental profits. When Snow White finally came to DVD 50 years after it's initial release, it made an absolute fortune (I think it immediately sold 30 million units).

And when a new format replaces DVD, then it will be sold in large quantities again (as will Toy Story). Good animation is evergreen, and Toy Story will likely be making coin for Disney 50 years from now just like Snow White did.

As for the ongoing Care Bears franchise, again, I'm not sure how its relative success hurts the argument that one can predict successful animated projects above a random rate. Does it weaken my argument that "high quality" increases the chance of success? I don't think so. Like, say The Rugrats Movie, it was a cheap movie with a built-in audience that became a modest success theatrically. But if you look at the real blockbusters, they're ALL high budget well-crafted projects. That's all I was saying.

Your Care Bear example just shows another avenue of avoiding randomness and being successful at picking profitable animated features: combine a built-in or niche audience with low production costs -- we can name more than a few animated films that meet that criteria, and did better than random returns.

Anonymous said...

When Snow White finally came to DVD 50 years after it's initial release, it made an absolute fortune (I think it immediately sold 30 million units).

Most if not all of the Disney features have done extremely well in VHS/DVD.

Name a live-action feature from 1937 that has sold 30 or even 3 million DVDs. There aren't any.

"Gone With the Wind" is the only live-action feature that has had the kind of sales and staying power that Disney animated features have.

Site Meter