As often happens in comment threads, the discussion for my post [A hard time to be an indie](http://johnaugust.com/archives/2009/hard-indie) focused less on the original article and more on the observations of a single commenter. In this case, [Rebecca](http://johnaugust.com/archives/2009/hard-indie#comment-171902):
> I’ve always wondered why the movie Lars and the Real Girl wasn’t released more widely. I only read rave reviews about it and everyone I know who saw it loved it. I thoroughly enjoyed it. They didn’t even release it widely enough to make a profit, the dumbasses.
> Here are the numbers:
> The movie had a $12,000,000 budget. [According to boxofficemojo.com](http://www.boxofficemojo.com/movies/?page=weekend&id=larsandtherealgirl.htm), it made $90,418 opening weekend in 7 theaters for a $12,916 average per theater.
> The Proposal opened this past weekend in 3056 theaters and grossed $33,627,598 for an average of $11,004 per theater.
> At 5 weeks, Lars and the Real Girl averaged $2,456 per theater after the number of theaters was reduced from its peak the week before.
> The latest Night at the Museum, just averaged $2,636 this week, its 5th.
> It looks to me like decent marketing in various markets, in conjunction with a much wider release, could have made this movie -– and everyone involved with it -– a LOT more money. Can you explain why it would not have made sense to release it into more than 321 theaters during its entire run? Other than thinking that challenging, quirky and maddening = noncommercial, I mean?
Apples, meet oranges.
Per-screen average is simply math: a given film’s box office divided by the number of screens it plays on. ((And even then, it’s a messy measurement. Particularly with wide releases, theaters can increase or decrease the number of screens devoted to a picture even over the course of a weekend, based on demand and sell-outs.)) As a pure number, it tells you nothing about the size of theater, the percentage of seats sold, or what would be typical for that theater on that night.
Bringing in $2,300 over a weekend might be a great haul at a tiny theater in Des Moines, but would be a disaster at Grauman’s Chinese.
The number is only useful when comparing movies in fairly similar situations. If The Happy Harpist made $44,000 at four theaters, and My Third Elbow made $10,000 at three, it’s fair to say that Harpist is outperforming Elbow with an $11,000 per-screen average.
But drill deeper, and you might find reasons why Harpist’s numbers are misleading. For example, it’s possible Harpist made $34,000 on one of its screens, and only $10,000 on the other three. Maybe it’s a hometown director, or other special circumstance. ((Or maybe it’s The Nines. We debuted on two screens — one in New York, one in LA. Two-thirds of our money came from LA’s NuArt.)) Take away that one theater, and Harpist and Elbow are now a dead heat.
More importantly, if you’re one of the low-performing theaters for Harpist, *your* per-screen average is only $3,333. You will make your decisions about whether to keep playing the movie based on that number. Never forget that distributors don’t ultimately decide which movies stick around in theaters; the exhibitors do. They look at their internal numbers to decide which movies will make them the most money.
**With a small number of screens, per-screen average is hugely affected by variations between individual venues.** The denominator — which screens, and where — matters a lot.
Conversely…
**With a big number of screens, per-screen average is relatively unaffected by variations between individual venues.** If you’re playing in 4,000+ screens, it doesn’t matter nearly as much which screens those are. You’re a wide release, playing at every other megaplex in the country. Distributors desperately scramble to get as many good screens as they can, simply so they can generate as much money as they can. Per-screen average is the last thing on their mind.
Some movies are able to successfully platform (expand) from a few screens to a lot. Juno, for example. But if you look at Juno’s [weekend boxoffice breakdown](http://www.boxofficemojo.com/movies/?page=weekend&id=juno.htm), you’ll see that it never came close to its opening weekend $59,124 per-screen average again. As it climbed to 2,000 screens, the per-screen average plummeted because the denominator had gotten so big. Trust me: Fox Searchlight didn’t care. They were too busy making gobs of money.
Same for The Proposal. Same for Night at the Museum 2. Unlike the makers of Lars and the Real Girl, who carefully selected each of the seven venues it debuted upon — like Goldilocks, not too big, not too small — the studios releasing blockbusters want as many seats as possible. They’re not looking to expand. They don’t need to nurture. They simply want the maximum amount of money, preferably in the shortest amount of time.
Rebecca points to the fifth weekend of Night at the Museum 2 and its $2,636 per-screen average. She conveniently omits that on that weekend it earned $7.8 million. **Money is money. Per-screen average is just a figure.**
Back to Lars
—–
While it’s absolutely fair to play Monday morning quarterback on a movie you love and believe could have made more money, the folks who released Lars and the Real Girl are not dumbasses. You can disagree with their marketing and perhaps their release date. I wouldn’t be surprised if the filmmakers feel disappointed. But they clearly tried to platform the movie much like Juno, and it didn’t work.
Courtesy Box Office Mojo, here are the [numbers for Juno](http://www.boxofficemojo.com/movies/?page=weekend&id=juno.htm):
And here’s [Lars](http://www.boxofficemojo.com/movies/?page=weekend&id=larsandtherealgirl.htm):
Both Juno and Lars started in three theaters, then expanded to 300 in their third week. But Juno far out-earned Lars at every step. By the time it went wide, Juno also had the advantage of the Christmas holiday.
As you’re looking at the Lars chart, rather than focusing on the per-screen average, look at the red numbers in the % Change column. Starting with Nov 2-4, it was making less each week. It was on a decline. The distributor couldn’t justify the millions of dollars it would take to expand the run when it was earning a fraction of that each week.
In the end, Lars and the Real Girl made just under $6 million domestically. Many indies would love to reach that number.
Could Lars have made more money? Perhaps with a different combination of marketing and luck. But per-screen average has nothing to do with it, and using that figure to compare it to wide releases is specious. Limited releases have high per-screen averages *because* they’re on so few screens, not despite it.
As it happens, all four of Stephenie Meyer’s Twilight books fall into this range, and have been here for quite a while. They’re priced at $5.50 (for the first two) and $9.99 (for the second two).