Why the Box Office Tracking numbers have been wrong so often this year, and why John Carter is really tracking at 40m, not 25m

For weeks now, Boxoffice.com has had John Carter opening at $25m, and yesterday the Los Angeles Times reported that a major tracking service had downgraded the projected opening from $30m to $25m.  So $25m is emerging as “the number”, according to the tracking services.  If this turns out to be true and John Carter opens at that low number, it will be a major disappointment for Disney and probably spell doom for any hope of a sequel for the sci-fi epic that Disney with the clear intention that it would spawn a franchise.  With the first three books already optioned, and a total of 11 books available – it’s clear what Disney hopes will happen.  But $25m opening weekend won’t get it done.  

So with the dismal figure of $25m being bandied about — does John Carter still have a realistic chance to open well enough so that, with legs and good overseas BO, a sequel can be generated?  Or is the writing already on the wall?  Our view is this: A disciplined look at the available data suggests that no, the writing is not yet on the wall, and in fact John Carter is likely tracking at about $40m if all factors are considered.

Let’s start by looking back a week.  The same tracking outfits that have John Carter at $25m had Lorax at a 40-50m opening.  It opened at $70m.  A few weeks back they had Safe House and The Vow at 20m — they opened at 40m.  Although it would be a stretch to say that the tracking has low for every major release this year — the general pattern that has emerged is just that.  The new films are frequently, if not consistently, outperforming the tracking estimates.   Not surprisingly — overall box office gross is up by 16% versus the same period a year ago, and there is nothing in the week-to-week trending that would suggest there is any softening as March 9 approaches.  So JC enters the marketplace at a time when tracking is frequently failing to accurately predict the opening — and when there is an overall robust market running 16% ahead of the same period last year, and 24% ahead in the most recent weekend.  Both of these factors suggest that JC may perform better than the tracking estimates.

But it’s not enough to just assume that the tracking estimates are low.   There is data available to anyone who cares to drill down into it which can help identify whether there is a likely flaw in the tracking figures, or not.   All one has to do is go to Box Office Mojo, open up an excel spreadsheet, and start drilling down into the data that is there — data that includes day by day box office gross for every major film; plus weekend by weekend; year against year; genre tracking; “similar movie” tracking, etc, etc.

Our little investigation started from the (reasonable, we think) assumption that if the tracking figures for JC are reasonably on target, then they should correlate with the overall weekend Top 12 Gross — or in other words, if you plug in the $25m JC tracking number and the tracking numbers for the other two new releases; then apply prudent and relatively easy-to-predict numbers for all the other films (keeping in mind that films already released have established patterns which make the prediction much easier) — then what should happen is that the number that the top 12 films add up to should be a number that makes sense for the overall weekend total — keeping in mind that aside from tracking individual films, an entire separate discipline is tracking the overall weekend gross (or Top 12 Gross, which is more manageable).  There should be “harmony” between the tracking for the new releases, and this overall Top 12 Total.

Is there?

We considered:

• “Tracking” as predicted by the tracking polls.
• Recent history of Tracking Polls vs Actuals
• Total Weekend BOG (or Top 12) for previous 10 weeks, and for the same period a year ago.
• Any “X Factors” that can be defined

New Films Opening March 9
First of all, the tracking estimates are, on average,  the following for the three new releases:

• John Carter – $25m
• Silent House – $11m
• A Thousand Words – $6m

The first question that comes to mind is — how do these numbers square with the macro picture of the overall Box Office Gross. In other words, if you plug those three numbers into a model of the overall box office gross, weekend by weekend and this year against last year — do the numbers make sense?

Well, we did that.  First of all it’s important to understand that generally speaking, the volatile portion of the box office are the New Releases — these are the films who’s outcome is hard to predict.  Once a film has been theaters for even one weekend, there are patterns that tell you what will happen next, once the first weekend has been established.  For example — an average second week drop off is 35-45% but it’s possible to estimate more closely than that by studying the drop-offs experienced by the film’s director in his/her previous movies, or earlier editions in the franchise.  In the case of Lorax, there is very strong evidence to suggest that the drop off will be right around 40%.

So right away, if you apply the 40% drop off to Lorax, and reasonable, easy to predict (because they are already in the marketplace) drop-offs for the other films already in release– and you apply the tracking numbers listed above for the new releases, the result is $117m for overall Top 12 Box Office Gross for the weekend of March 9.   That would be on the heels of a weekend–March 2–in which Top 12 Box Office Gross was $153m.  That means a weekend-to-weekend drop of 24%  How often does that happen?  Answer, not very often.

So right away, there’s a disparity between the predicted overall BOG using the tracking estimates, and the predicted overall BOG just using a Macro model of the overall trends.

Looking a little deeper — if you look at the last five years and compare the last weekend of Feb, with the second weekend of March, an interesting pattern emerges.  And by the way — why choose the last week of Feb , then skip a week?  Answer: Because the last week of Feb is usually the end of the “doldrums” with a major film typically released on the first weekend in March — creating substantial volatility for the first weekend of March.  The patterns are quite a bit more consistent when you apply this metric, and thus more meaningful for projections.

So if you look at this–in four of the last five years the 2nd Weekend in March has been up an average of about 25% over the last weekend in Feb.   Now remember — using the “tracking” model, we get 117m for March 9 weekend.  Using this model we get $141M.  Meaning — if you simply apply the average increase from Last Weekend in Feb to Second Weekend in March — you get a Top 12 BOG of 141M, not $117m.

Are there other reasons we should consider the overall figure of $117m to be suspect?

Well, for starters, last year the Top 12 for the same weekend was $114m and this year is running 16% ahead of last year, so using that metric the projection for March 9 would be 135M.   But remember — last year in this frame Disney released the mega-disaster Mars Needs Moms and the overall box office was still at 114m.  Does anyone seriously think that Disney’s entry this year will just do $6m?

Okay — so just using a couple of fairly simple tools as described above, it seems highly unlikely that overall Top 12 BOG will end up at 117m as predicted by the “tracking” model.  So let’s drill down a little more deeply.  The following chart contains the actuals for March 2 weekend for all of the Top 12 films; then three projection models — each of which includes “plenty of room for Lorax” while also taking into consideration the overall Top 12 figures.

• Per Tracking:  This model is what we can expect if the tracking is correct.  It uses the average figures for the new releases “as per tracking” estimates; it applies a 40% drop off to Lorax which is strongly indicated based on the com parables; and it applies the same drop-offs to the other films that have been in release multi-weeks that they had last week.  (The one wildcard is The Artist who is in the midst of its Academy Award ‘bump’, but the numbers are still very small and not a major impact.)
• Realistic Minimum:  This model simply applies the benefit of this year’s better Box Office performance as measured by the year to year comparison, and allocates the increase to the new films in proportion to their tracking estimates.  In other words, the tracking has JC getting 57% of the “new film” share of the box office, so this share is retained.
• Probable:  This one takes into consideration the year to year comparison, and the “Last Week Feb/2nd Week March” comparison, and comes up with what would seem to be a probable outcome.

So is that the bottom line?

Not quite.

What about “X Factors” that might affect the outcome further?  For example, is there a substantial group of potential viewers of John Carter who are not part of the “BOG Economy” of this year to date?  The argument could be made that yes, there are at least $2-3M worth of tickets to be sold to Boomer Burroughs Fans who don’t show up in the tracking; haven’t been to a movie this year; but will go to this one and bring kids and grandkids.  The ERB novels were hugely popular and widely available in the 60’s and 70’s — that’s how Cameron, Lucas, Spielberg, Carl Sagan, and the rest of the famous ERB fans got their hands on them.  So there’s that.

Secondly, there is the Disney Factor.  Disney films frequently beat their tracking estimates even when there are no macro factors driving box office up like there are now.  Disney is more of a brand than any other studio; and while there are plenty of young adults for whom Disney is not a positive (a factor already recognized in the weak tracking), there are plenty of families who may not be showing up in the tracking.

Finally, there is the generally positive word of mouth that is only gaining momentum in the final days before the film is released and is thus not captured by the tracking from last weekend.  This may help opening day — and it may help the Saturday/Sunday numbers.

Where does that leave it?

Conservatively, $40m.

Probably $45m.

Optimistically: $50m.

What Do These Figures Mean for the Prospects of a Sequel?

(Stay tuned….we’ll talk about that separately.)