Tuesday 12 March 2013

How do I interpret midnight box office data?

Midnight box office is the earliest real indicator of whether an audience is actually going to turn out to see a movie.  Sure, we can get early clues from pre-release indicators like survey-based tracking, advanced sales and social networking data from sources like Flixster and Twitter.  But midnight box office is where the rubber meets the road.


Interpreting midnight box office is more of an art than a science, but having the numbers and comparing them against other similar movies is still helpful in guiding your last-minute guesses and predictions.  This column will look at some of the factors to consider in interpreting midnight box office.  The data used here can be accessed from the link in this column.

The midnight movie business is new and growing

A decade ago, only the very biggest tentpoles - like the Star Wars, Spider-Man and Lord of the Rings movies - would get midnight releases.  But as midnight showings got bigger and bigger, more and more movies started getting midnight showings.  Between 1999 and 2005, eight movies got midnight releases.  In the last twelve months as of writing, there have been over thirty.  For movies that appeal to teenagers, fanboys and fangirls and horror audiences, midnight screenings are now expected.

The increased popularity of midnight screenings also mean that you have to be careful in using midnight box office data from movies releases more than a couple of years ago.  To take an obvious example, here are the midnight numbers for the last five Harry Potter movies, and their box office totals for the next three days of release (not including midnights):


Movie
Release Date
Midnight Box Office ($M)
Next Three Days Box Office ($M)
Harry Potter 4 - Goblet of Fire
18-Nov-05
6.3
96.4
Harry Potter 5 - Order of the Phoenix
11-Jul-07
12
76.4
Harry Potter 6 - Half-Blood Prince
15-Jul-09
22.2
84.8
Harry Potter 7 - Deathly Hallows 1
19-Nov-10
24
101.0
Harry Potter 8 - Deathly Hallows 2
15-Jul-11
43.5
125.7

Harry Potter 5 nearly doubled the midnight take of Harry Potter 4, but took about 20 per cent less box office over the next three days.  Similar result with Harry Potter 6 - much more midnight box office, 12 per cent lower over the 3-day.  Of course, Harry Potters 5 and 6 were Wednesday releases which partly explains the lower three-day take.  However, Harry Potter 7 was also a Friday release and while it took four times the midnight gross of Harry Potter 4, it earned only marginally more over the next three days.

The most reliable midnight audiences are fangirls and fanboys aged under 25

Which audience segments are most likely to go to midnight screenings?  Not adults with jobs or other responsibilities, because they have to get up early the next morning.  Not young children, because it's way past their bedtime.  No, the audience segments that make up the bulk of midnight audiences are teenagers, other people aged under 25 and fanboys and fangirls who are highly motivated to see the movie with (or before) other enthusiastic fans.

To support this argument, here are the top ten midnight box office earners of all time, as of writing:


Movie
Midnight Box Office ($M)
Harry Potter 8 - Deathly Hallows 2
43.50
Batman (Nolan) 3 - Dark Knight Rises
30.60
Twilight 5 - Breaking Dawn 2
30.40
Twilight 4 - Breaking Dawn 1
30.25
Twilight 3 - Eclipse
30.10
Twilight 2 - New Moon
26.30
Harry Potter 7 - Deathly Hallows 1
24.00
Harry Potter 6 - Half-Blood Prince
22.20
Hunger Games 1
19.70
Avengers (Marvel)
18.70

You don't have to look very closely to see that they have a lot in common.  There are four Twilight movies on the list, and three Harry Potters as well as the first Hunger Games movie.  All of these are all highly-hyped movies aimed squarely at teenage/under 25 fangirl/fanboy audiences.  The other two ate the latest Batman movie and the first Avengers movie, which are also highly fanboy/fangirl driven, and don't exactly ignore teenagers and other under 25s either.

There is a relationship between midnight box office and weekend box office

There certainly isn't a straight line relationship between midnight box office and total box office for the weekend.  The movie with the biggest opening weekend ever - The Avengers - is ranked #10 on the list of biggest midnight movies.  But there's still a relationship.

This chart has data for midnight box office on the X-axis, and what we will call the next day multiplier on the Y-axis.  The next day multiplier is calculated by taking the box office for the next day (not including midnights) and dividing it by the midnight box office number.  Looking at the relationship between the two, the trendline has an R2 of over 0.75, indicating a pretty strong correlation between a movie's midnight box office and its earnings for the next day.  (Off-column, I've done a chart comparing midnights against three day totals, and it produces a similar result.)




This includes all the midnight box office I have data for, including the older releases.  Not all of the next days in the charts are Fridays, because I have included data from mid-week midnights, and some box office numbers include pre-midnight screenings.

Those qualifications aside, this is simple! To predict how well a movie will do from the midnight numbers, all we have to do it plug the box office figure into the formula on the chart, right?  Well, not quite.

Look closer at the chart. The trendline does a good job of predicting where the box office will end up when a movie earns a lot of midnight box office.  But the lower the midnight box office, the greater the variation.  Some movies with a midnight take of $500K have a next day multiplier of around 5, others with the same midnight numbers have a next day multiplier of 20 or 25.

So how do you tell if $500K in midnight box office is a good number or a bad one?

The younger the audience, the bigger the front-loading

Let's look more closely at some of the data points on the chart above.  For this exercise, I stripped out every movie released before 2008 and every movie that earned over $2 million in midnights, as well as a few other movies with odd release patterns or unsubstantiated midnight numbers.  Of the movies remaining in my data set, here are the ten with the highest next day multipliers.


Movie
Release Date
Midnight Box Office ($M)
Next Day Box Office ($M - without Midnights)
Next Day Multiplier
Grown Ups
06-Aug-10
0.55
13.9
25.2
Safe House
10-Feb-12
0.54
13.1
24.3
Super 8
10-Jun-11
0.5
11.6
23.3
Other Guys
25-Jun-10
0.55
12.6
22.9
2012
13-Nov-09
1
22.4
22.4
Salt
23-Jul-10
0.55
12.0
21.8
Think Like a Man
20-Apr-12
0.55
11.6
21.0
The Vow
10-Feb-12
0.7
14.6
20.8
Battleship
18-May-12
0.42
8.2
19.6
Green Hornet
14-Jan-11
0.55
10.5
19.2

The first thing you notice is that these movies have adults as a key section of their audience.  Sure, a lot of under 25s like Will Ferrell and Adam Sandler comedies, Denzel Washington and Angelina Jolie action vehicles, movies with lots of explosions like 2012 or Battleship, or movies that sound like they were written by Nicholas Sparks.  But these movies also have a lot of appeal to adults, who are happy to wait for the weekend to see them.

Here are the movies from the same data set with the ten lowest next day multipliers:

Movie
Release Date
Midnight Box Office ($M)
Next Day Box Office ($M - without Midnights)
Next Day Multiplier
Saw 7
29-Oct-10
1.7
7.3
4.3
Chernobyl Diaries
25-May-12
0.55
3.0
5.4
Dictator
16-May-12
0.65
3.5
5.4
Project X
02-Mar-12
1.2
6.9
5.8
Sinister
14-Oct-12
1
6.4
6.4
Inglourious Basterds
21-Aug-09
1.9
12.5
6.6
Star Wars 1 – 3D Re-Release
10-Feb-12
1.1
7.5
6.8
Act of Valor
24-Feb-12
1.1
7.9
7.2
Devil Inside
06-Jan-12
2
14.8
7.4
District 9
14-Aug-09
1.6
12.6
7.9

Males under 25s are the core audience for most of these movies - and four of the five on the list with the lowest next day multipliers appeal to pretty much nobody except for males under 25s or horror fans.  While others like The Dictator, Inglourious Basterds and District 9 have broader appeal, there's a definite fanboy audience for each movie.

So, when the next midnight box office info is posted on the Movies board and you are trying to figure out what these numbers mean for the moviestock adjust:
  • midnight screenings have really only taken off since 2012 or so, so recent movies make better comparisons than older ones
  • midnight box office is driven by under 25s and motivated fanboys and fangirls.  A movie that appeals to both can achieve a huge midnight number
  • the bigger the midnight box office, the easier it is to predict the box office for the next day and for the opening weekend
  • for movies with a small midnight number, it pays to compare movies with similar audiences.  A small midnight number does not necessarily mean a small opening weekend, particularly if the movie has a lot of appeal to adult audiences, who do not regularly attend midnight screenings, while a slightly larger midnight number can still mean a weak box office if the movie only appeals to a single audience segment.

No comments:

Post a Comment