Does a bigger film production budget result in more ticket sales?

If you take a stroll down a list of the most expensive films of all time, you’ll notice that most of the films are from the past 15 years. Every year, more and more money is being poured into producing larger and grander films, likely with the hope that a bigger film production budget will result in more movie ticket sales. For the chart below, I analyzed the production budget and ticket sales data of 11,706 films (courtesy of Box Office Mojo) to see whether that assumption held up.

film-sales-vs-budget-correlation

(Note: All dollar values have been adjusted to 2014 dollars.)

It’s fun to note some of the outlier cases in the above chart:

  • Midget Zombie Takeover is the most unremarkable film on this chart (in terms of budget and ticket sales), with a measly budget of $2,043 and ticket sales totaling $10,653.
  • Tarnation is the biggest underdog success story with ticket sales totaling $721,546 coming from a budget of only $277.
  • Zyzzyx Road is the biggest failure with a paltry $35 in ticket sales despite a $2.36 million production budget, although the poor ticket sales were due to a deliberate effort by the producer.

Amusing outliers aside, the above chart is likely misleading because the few outlier cases in the lower budget and ticket sales ranges are strongly affecting the linear regression. For that reason, I focused the analysis only on big-budget films with at least $1 million in both budget and ticket sales.

film-sales-vs-budget-correlation-zoomed

Although there’s a weak correlation between film production budget and ticket sales (R^2 = 0.32), it’s fairly clear that just pouring money into a film’s production budget to hire high-profile actors, add more CGI, etc. doesn’t mean that the film will sell more tickets. In fact, if we compare the linear regression (blue line) to the line of parity (black line), the more that’s spent on film production, the less likely the film will end up making that investment back in ticket sales.

Regardless, it’s interesting to see that there’s even a weak correlation between a film’s budget and it’s performance in the box office. Only 4 films with a $100M budget ever made less than $10M in ticket sales, whereas only 6 films with a budget less than $10M ever made more than $100M in ticket sales. It seems that the film industry is like the stock market: You have to spend money to make money. And if you want to make the real big money, you’re gambling with an awful lot of benjamins.

Dr. Randy Olson is a Senior Data Scientist at the University of Pennsylvania, where he develops state-of-the-art machine learning algorithms with a focus on biomedical applications.

Posted in analysis, data visualization Tagged with: , ,
  • Pingback: The biggest box office booms and busts since 1982 | Randal S. Olson()

  • Fascinating. I am wondering what this might look like if you looked at box office gross to deal with variances arising out of the infamous movie studio accounting systems, which have been known to dramatically skew reported profits in some instances.

    Love the blog, by the way.

  • Zed

    Do the trend lines do anything interesting when split between Big 6 movies and others/indies?

    • Interesting idea! I’ll add this to the pile of things to look at when I revisit this data set. 🙂

  • Ryan

    I would say there is a pretty strong correlation. R^2 tells you that 32% of the variability in log(profits) can be explained by log(budget). The correlation coefficient is roughly .57. I would say that is a pretty good correlation.

  • MikeW

    A) Interesting.
    B) I think your labels are misleading. You’ve labeled your axes as log(thing), but you didn’t actually plot log(thing). You plotted thing on a log scale. If the log of the production budget were at least $1m, that would be one absolutely gigantic budget.

    • True. Labeling it as log(X) just seemed to be the most concise way to indicate that the axes were on log scale. 🙂

  • Ryan

    Maybe I’m interpreting the graph incorrectly, but I don’t think it is true to say “the more that’s spent on film production, the less likely the film will end up making that investment back in ticket sales.” The model estimates that there will be, on average, profits of $100 million when the production budget is $110 million. They made the $110 million back plus another $100 million. Wouldn’t a profit of $0 mean the investment was exactly met? Any additional revenue will contribute to a profit above $0, which is shown in the graph.

    • Jean-Paul

      I think Randy may have confused ‘profits’ with ‘grosses’ a little. Box-office mojo only provides box-office grosses. And generally, as Box-office mojo’s specifies, you have to roughly cut down the grosses in half, so one part goes to the theaters, the rest goes to the distributor (distributor = the studio when it’s one of the majors’ production, it’s not the case when dealing with independent films, but there ‘s still a big and rougly equal chunk going to the theater).

      So, if we are really talking about profits here as “movie box-office profits” from the chart suggests, then what the second chart shows is that all the movies of the selection were profitable ! (they made more money than they cost). And that, on average, the more they cost, the more profits they made; which would definitively prove that the more you spend on a movie the more luckily you’ll benefit from your investment.

      Now, if we are dealing with mere box-office grosses (which would be the equivalent of revenue), as “Although there’s a weak correlation between film production budget and ticket sales” suggests from the commentary of the second chart ( ticket sales is another term for box-office grosses, but means something different than profits), all the second chart shows, is that, globally, the more you spend, as a studio, the higher the ticket sales/the box-office grosses; which still doesnt say much about Return on Investment / Profitability of your investment in your film. However “the more that’s spent on film production, the less likely the film will end up making that investment back in ticket sales” remains true according to second chart and since Randy is talking here about tickets sales, not profits.

      I’m only a student in the film industry, and english is not my first language, so please correct me if I’m wrong, because I find this article in anycase very interesting, thanks in advance.

      • Jean-Paul is correct that I mislabeled the axes. They should read “grosses”, not “profits.”

  • Saulih

    Would be interesting to compare whether the payoff for film budget expenditure is dependent on the CGI expense, director, producer, or actor. Is having a ‘top shelf director’ a better choice for turning a profit, or spending money on CGI? Does a seasoned actor trump the director?

  • NZ Finance guy

    Great read – and this prompted me to thinking about all the comic-based movies we’ve been getting lately. I mainly focused on the Marvel vs DC competition, but came to similar conclusions re budget vs gross takings over this smaller dataset.

    http://nzfinanceguy.blogspot.co.nz/2016/08/having-stumbled-across-great-blog-of.html