Comic Fodder

Who Watches the Statistical Regression?

"The movie is a hit!" "It failed expectations!" "We know how much/little it will make!"

There has been a firestorm of articles about the newest comic book movie, and as usual, any time the concentration level of the media reaches a certain level to attract what is commonly referred to as the mainstream media, well, the errors, they will begin. Let me make my own humble entry to help the poor reader make sense out of all of the contradictory stories that have been coming out over the past week.

Let's start with the success/failure of the movie. In an interesting way, movies themselves are received in much the same manner that earnings reports are received by the investment community: you can give them great numbers, but if you fail to meet expectations, then you're somehow a dud. Case in point: Superman Returns made $200 million domestic, and pulled in almost $400 million worldwide, but there will be no sequel anytime soon. Why? Because it failed expectations. It made nowhere near what it was "supposed to" for the first day, for its opening weekend, nor for its entire run. This was "supposed" to be a $400 million movie for domestic receipts alone! That's the way the most recognizable comic property was treated, with comparative examples like Spider-Man and X-Men setting the precedent. How would the mavens treat Watchmen, one of the most well-known comics in history, but still relatively unknown to the rest of the world at large?

The people in the know watched the journey of the film and came close. The box office gross for the first weekend was $55.7 million, and it was the #1 movie of the week. Not bad, right? Too bad, Yahoo decided it "fell short of expectations." The actual company that made the movie said it they were happy, but whether you believe them or not really depends upon your existing bias; if you thought it was the right number, you would think Warner Brothers was being honest. If you suspected it was a disappointment, you would accuse them of spin. Evidently a lot of the people who spend their time predicting opening weekend numbers wanted to see $60 million or more, so they were "disappointed." For some reason, they expected 300 numbers. It's like they don't even follow the industry for which they are prognosticating.

300 was an 'R'-rated film which opened at a slow time during the movie-going seasonal cycle, and it opened to a record-shattering $71 million weekend. Somehow this became the new standard by which to measure films in a similar genre. And I don't just mean 'standard' in the sense of the noun, in an approved method for comparison, I mean for the adjective too: this became the new, default, standard measurement. The critics and forecasters all seem to have conveniently forgotten that the opening weekend of 300 was a big surprise. Almost nobody saw it coming. How in the world did a surprise hit end up subverting the existing baseline of box office forecasting?

They also tend to want the movie to have the same gross as the director's previous movie, as if this had any bearing on the subject. It can have a bearing, for similar types of movies, and I know for the normal predictors, they might have thought 300 was the same as Watchmen, basically. Trust me, it isn't. Having a director or star is no guarantee that the movie will approach the gross of another movie by the same director/star. Even big box office draws like Brad Pitt may choose an indie project for their next film, and have it pull in less than $3 million. Six months later, he'll be in another $100 million blockbuster. Even Steven Spielberg, whose name will always be in the top ten list for names that can pull in crowds by name recognition alone, can't guarantee a consistent gross for movies. His Munich project only pulled in $47 million. The Terminal only $77 million, a pittance compared to his more famous successes. Having Zack Snyder as the director for his third movie should in no way have created a forecasting close to that of his previous movie. The comparisons drawn for either genre reasons or director reasons are entirely baseless in this case.

What factors are relevant? One of the fundamentals of reality, for starters: running time. Two hours and forty minutes plus previews, folks. That limits how many showings you can have for the movie. The rating, for another. With the average member of the public thinking at first this was a normal superhero movie, I came across a ton of genuine startlement at the idea that this was an 'R'-rated movie, and not something to which you should bring the kids. Did these forecasters take the running time into account? I think not. Did they understand that 300 was not viewed by the general public as a "superhero" movie, but something more like the action-oriented Troy? I think not.

How did HSX do? HSX is a long-running game that takes movies and stars and turns it into a stock market. The closing prediction for the opening weekend of the movie turned out to be somewhere around $62 million, in range with most of the predictions for the rest of the forecasting community. Does $55.7 really make such a big difference that the headline of the Yahoo article has to imply that the movie is a failure?

The sheer endurance of this project as an idea, combined with its tortuous route into becoming a film in the first place makes it as much of a miracle as any other project, that it could actually succeed in making it to the big screen in the first place. For a series that was not a household name like Superman, Batman, Spider-Man, or even Blade or Catwoman for that matter, to pull in an opening weekend like this, outside of the heavy summer or holiday attendance season, is impressive. Keep in mind that the star characters were all (mostly) relatively obscure, and grant that the subject matter is more complex than the average movie plot presents, and Watchmen's opening weekend gross takes on a new light. Do not be fooled by people trying to come up with a snappy headline: the first weekend was a definite success, and with the groundwork already laid for two successive versions of DVD release, each with additional material, PLUS the fact that it was #1 in the rest of the world, with a gross of $27.5 million, to make for a grand total of $83.2 million for its worldwide release, you have a blockbuster on your hands.

Now let's turn to forecasting for its entire theatrical run in the United States. Chadwick Matlin and Chris Wilson claim to have already figured it out. The duo wanted to see if critics' reviews had any measurable effect upon a movie not for the opening weekend, but for the time after that. Do the critics matter? Can their wrath bring down the total box office gross of an average theatre run? These guys take all the archive data they can get their hands on, crunch some numbers, and spit out a statistically insignificant answer that reaffirms common sense: if the critics rave about it, it does slightly better.

They do single out a stronger correlation for action movies, and try to argue that the opening number and positive critical reviews do help their predictive analysis, and their final conclusion is that the opening weekend of Watchmen will make up 32.7% of its total box office gross. So the end number for Watchmen will be $170 million. The adjusting calculation for Watchmen on the HSX site puts it at $150 million for the first four weeks of its release, which happens to be in line with this pair's prediction, since the movie will probably be out in theatres for a while, and still be showing on a handful of screens 30 weeks from now. But will this statistical number reflect reality? Yet again, I think not.

Some forecasters have seen an increase in ticket sales industry-wide and concluded hit movies will do well this year. Drawing solely upon my own experience following predictions, and running an account on HSX for several years (I bought shares of Watchmen at $85, so I ended up with several million dollars in fake money for profits when I cashed out), my feeling is that this movie will not have the average legs of your normal blockbuster. Part of it is still the physical restrictions of running time, part of it is the hype and overloading of the opening weekend. Part of it is my sense of the non-comics audience, many of whom did not completely follow the plot. All of these inputs congeal to form my gut instinct, leaving me with a non-statistical hunch that nevertheless, I could use to calculate a number. I predict for the first four weeks of release, the movie will pull in around $125 million, give or take 5 mil. For its entire theatrical run, it will rake in $150 million. That's $20 million less than the number crunchers and their multiple decades of movie archive data, statistical numbers, regression analysis, and fun interactive cinematic tools (which I recommend you play around with if you get a chance).

The conclusion of the article is actually a little confusing. They mention poor critical reviews may contribute to a flameout. Does this mean the existing poor reviews of Watchmen have already been factored in to account for their 32.7% number? Or do they mean to say that their starting number for the opening weekend is 32.7% of the total, and the grand total will go down from there if there actually exists a positive correlation between blockbusters and critics' reviews? They do not specify in the article itself.

An e-mail to the authors yielded clarity, and they do tend to think that the starting point is 32.7%, and that amount of front-loading alone is a potential signal of a flameout. So poor critical reviews may bite into the grand total, but that still means they think it will end up near $170 million. I will follow up to see how close that "flameout" might really end up, as compared to my $150 million estimation. So, I guess I'm saying that… I will keep watching.
Tpull is Travis Pullen. He started reading comics at 5 years old, and he can't seem to stop.

Im thinking and have been thinking more around 105 million for 4 weeks.

-- Posted by: jeff at March 12, 2009 7:35 AM

Well, this doesn't bode well for non-Fanboy word of mouth:

My guess is that if more stories of that sort appear, that's not going to help anyone out.

The bottom line is that the studio had no idea what it was dealing with. They advertised an action movie. Aside from a very limited portion of the audience, people become uncomfortable when their expectations of a movie are defied.

Plus, no matter how many times you run a TV ad, people still have no idea what a "Watchmen" is, or why they should see it.

I think it'll do beautifully on DVD and Blu-Ray.

And, no, I still haven't seen it.

-- Posted by: Ryan at March 12, 2009 1:38 PM

Ryan hit the nail - without name recognition there was a need to tell the audience who the Watchmen are and that did not occur. Were the Watchmen the heroes? the villains? the audience?

And the running time is another. That's the difference between 8 and 10 showings in one day on one screens. That's a 25% increase, which applied to the box office would raise it to the expectation of $70M.

-- Posted by: David at March 13, 2009 11:19 AM

I'm interested in seeing who has the more accurate prognostication for Watchmen's ultimate earnings. I suspect it will earn its budget back easily, but will it make a profit? That's a bit more doubtful.

Generally speaking, an average film needs to earn 2.5 times its budget back in order to be considered a success. Blockbusters are a little different, because of advertising costs. 300 had a significantly smaller ad budget than Watchmen did, so it probably needed to only earn twice its budget back. 300 actually earned a bit more than 5 times its budget, which puts it into "smash hit" territory.

By contrast, one of the main reasons why Superman Returns was considered a failure by many despite earning a tad more than $390 million worldwide, is that its production costs were staggering. The "official" number was $270 million, which is bad enough, but knowledgeable insiders have pegged it near $300 million. (If one were to be mean-spirited and add in all the other money spent on the project over the years, the cost approaches $500 million. How a company spends nearly $200 million without a movie to show for it is a mystery to me.)

Start with the official budget, factor in the $42 million ad campaign and $10 million for prints for Superman Returns, and it needed to earn a bare minimum of $320 million just to cover its costs. Use the actual number and factor in all those things no one ever thinks about -- translating it into other languages, etc. -- and now you're talking $360 million. Just to cover the cost of making it.

At the end of the day, the theatrical release of Superman Returns lost a ton of money for the studio and, even worse, gave the franchise a black eye, reducing its value in the marketplace. SR earned a further $82 million on DVD and $23 million for cable rights, which is still nowhere near what it needed to truly be considered a success.

-- Posted by: Trike at March 14, 2009 12:04 AM

You could do a Chi Square test and determine the independence between revenue and cost of all the top ten films within a 5 year period. Testing would determine if in fact more money spent equals consistently higher returns. Since Superman was the most expensive movie in 2006, proving this would also essentially prove or show what expectations might have actually been put on it.

-- Posted by: Peppino at March 21, 2009 12:40 AM