Jump to content

The Panda

The Panda's Extremely Poor Simple Linear Regression Analysis: Do YouTube Trailer Views Correlate With Higher Box Office Grosses?

Recommended Posts

Here was my methodology (just so you know, it's rushed and shit, but it gives a nice picture)

 

I took the top 50 grossing movies from 2016, and I made a table of their domestic grosses, their domestic grosses and the YouTube views from their highest viewed trailers.  I then made scatteplots relating Domestic Gross (Y-Axis) to YouTube Views (X-Axis), and opening weekend gross (Y-Axis) to YouTube Views (X-Axis).  I added trendlines to each, and I had excel run a simple linear regression for Domestic Gross (Y-Axis) and YouTube Views (X-Axis), and Opening Weekend (Y-Axis) and YouTube Views (X-Axis).  

 

Here is what I found, please take it with a grain of salt because there is a lot I didn't account for.

 

First, here are my scatterplots (I didn't make them fancy or anything, default excel format with a little cleaning)

 

i8QniLW.jpg

 

tDblKsM.jpg

 

Next here are my linear regression results:

Domestic - Trailer Views Test

Multiple R 0.56

R-Square 0.31

Adjusted R-Square 0.30

Standard Error 101787510.8

Observations: 49

 

Intercept

Coefficient: 87,638,213.07

Standard Error: 22,095,204.1

p-value: 0.000247618***

 

YouTube Views

Coefficient: 3.13

Standard Error: 0.67

p-value: 2.84E-05***

 

OW - Trailer Views Test

Multiple R 0.58

R-Square 0.34

Adjusted R-Square 0.32

Standard Error 34944778.77

Observations 49

 

Intercept

Coefficient: 25,167,190.43

Standard Error: 7,585,528.06

p-value: 0.001757171**

 

YouTube Views

Coefficient: 1.13

Standard Error: 0.23

p-value: 1.24E-05***

 

 

My Quick Analysis About This:

 

To try to put these into terms that don't require you to know anything about statistics, this test would indicate that YouTube Trailer views are in fact good predictors of Box Office Gross, or at least correlate well with them.  Not only do we know that more views on a YouTube trailer correlates with higher BoxOffice growth, we can assign a value to it.

 

Every trailer view that a movie has should correlate to an increase of its opening weekend gross by about $1.13, assuming that the movie has a base level opening weekend of 25.2m.  Obviously, this isn't a perfect fit, especially for your few large opening movies with many views, but on average this fits.  Now, you also have to take into account the 7.6m standard error in the model, and that the model isn't taking into account any movie that wasn't in the top 50 for the year.  There are also more factors to take into account, such as Sing's trailer for some reason having over 100m skewing the model, and obviously that movie didn't open close to 100m.  

 

Every trailer view that a movie has should correlate to an increase of its domestic gross by about $3.13, assuming the movie has a base level domestic gross of about 87.6m.  As with the Opening Weekend, obviously not every movie is going to have that base for its domestic gross (as this model is only looking at the higher grossers of the year), but if we continued to add more data to the model I'd suspect that the results of how much you account YouTube views for would remain fairly constant to what it is in the top 50.  

 

The model was a slightly better fit to opening weekends rather than domestic gross, and that would make sense as WoM becomes an extra unaccounted for variable after the OW.  

 

Another neat observation

Hidden Figures was the only movie to gross over 100m that didn't have at 10m trailer views.  Also, your comic book movies tended to have more trailer views than movies that grosses similarly to them, while your more adult oriented films and animated films tended to have less compared to movies in a similar gross range to them.

 

Anyways, next time you're making your predictions for the Summer game or trying to decide whether or not to be IN or OUT of a risky club, remember trailer views are a strong indicator of Box Office Gross.  While, it is not a definite that a movie with more trailer views will gross higher than the one with less, on average it should.  You can also assume that the opening weekend of a movie should, in almost every case, at the very least equal the amount of YT trailer views it has, if not greatly exceed it (by anywhere from 2-4xish as much).

 

Also, if a movie is a big tentpole movie (and it's safe to assume an OW of at least 25.2m), take the number of trailer views it has, multiply it by 1.13 and add the answer to 25.2m.  That should get you somewhere in the ballpark of the OW on average, however take into consideration I am looking at Trailer views after the movie had opened, plus the model doesn't work as well for your big outlier openers, so you could very easily be undershooting the OW by quite a bit.

 

Hope you find this quick analysis useful for your box office predicting!  I may do another one of these looking at Tomatometers, but that'll require a different approach than what I did for this one.

 

  • Like 24
Link to comment
Share on other sites





4 minutes ago, WrathOfHan said:

And people say I base my predictions off YT views too much :redcapes: 

 

Even with this, I'd still be hesitant to base your predictions on YT views alone.  But bottom line is they are significant indicators of BO gross.

  • Like 4
Link to comment
Share on other sites



10 minutes ago, The Panda said:

 

Anyways, next time you're making your predictions for the Summer game or trying to decide whether or not to be IN or OUT of a risky club, remember trailer views are a strong indicator of Box Office Gross.  While, it is not a definite that a movie with more trailer views will gross higher than the one with less, on average it should.  You can also assume that the opening weekend of a movie should, in almost every case, at the very least equal the amount of YT trailer views it has, if not greatly exceed it (by anywhere from 2-4xish as much).

 

Also, if a movie is a big tentpole movie (and it's safe to assume an OW of at least 25.2m), take the number of trailer views it has, multiply it by 1.13 and add the answer to 25.2m.  That should get you somewhere in the ballpark of the OW on average, however take into consideration I am looking at Trailer views after the movie had opened, plus the model doesn't work as well for your big outlier openers, so you could very easily be undershooting the OW by quite a bit.

 

 

So you're saying that Valerian will open to 37M at least, but possibly more as Youtube views continue to increase until its opening right? It should get to at least 40M OW by the time it comes out basing it on your model which would indicate I predicted it perfectly on my club. I'm glad to be proven right.

 

Spoiler

 

:ph34r:

Edited by Arlborn
Link to comment
Share on other sites

1 minute ago, MrPink said:

Did you use trailer views as a single variable? I'm just curious if you factored other variables such as a franchise entry, reviews, etc

 

Yeah I was only measuring YouTube trailer views and DOM Gross/OW.

 

Sometime in the future I may try and do a better test with more movies (top 200), and add in a few extra variables like review, whether it's a franchise movie or not, whether it has Oscar nods or not, etc.

 

This analysis was pretty basic.

  • Like 2
Link to comment
Share on other sites



Just now, Arlborn said:

So you're saying that Valerian will open to 37M at least, but possibly more as Youtube views continue to increase until its opening right? It should get to at least 40M OW by the time it comes out basing it on your model which would indicate I predicted it perfectly on my club. I'm glad to be proven right.

 

  Reveal hidden contents

 

 

Maybe, maybe not.  I wouldn't expect every movie to match up to perfectly, it's a rather rough model.

Link to comment
Share on other sites



Interesting. I do think YT views are generally a good barometer for atleast the OW. However, specific films like comic-book, fan driven properties or franchises are heavily favored.

I think on your next study using Rotten Tomatoes, you find that it's probably more important tool to use for adult films and other original premises. 

  • Like 1
Link to comment
Share on other sites



3 minutes ago, The Panda said:

 

Even with this, I'd still be hesitant to base your predictions on YT views alone.  But bottom line is they are significant indicators of BO gross.

 

One metric I actually look at with regards to YT views is not just the views, but the drop off in views between the teaser and the trailer. If there is less than a 50% drop off, then it is fine most of the time, a bigger dropoff indicates a lack of interest in even following the trailer campaign post the initial teaser. I have also found that YT views are more for OW than they are for overall gross, YT views will almost correspond exactly to OW. Around 2 weeks or so ago, I was bored and actually calculated the view counts for each summer movie (not all uploads, just ones with more than 2M Vews). This is the list I had 2 weeks ago

 

GOTG 2 - 65M

King Arthur - 36M

Alien - 28M

Pirates - 46M

Baywatch - 30M

Wonder Woman - 58M

Mummy - 45M

Cars 3 - 40M

TF5 - 32M

DM3 - 52M

Spider-Man - 58M

Apes - 28M

Dunkirk - 28M

Valerian - 16M

 

Transformers 5 looks pretty awful in terms of trailer views to be honest. Might end up being the biggest domestic casualty of the year.

  • Like 7
Link to comment
Share on other sites





1 minute ago, grim22 said:

 

One metric I actually look at with regards to YT views is not just the views, but the drop off in views between the teaser and the trailer. If there is less than a 50% drop off, then it is fine most of the time, a bigger dropoff indicates a lack of interest in even following the trailer campaign post the initial teaser. I have also found that YT views are more for OW than they are for overall gross, YT views will almost correspond exactly to OW. Around 2 weeks or so ago, I was bored and actually calculated the view counts for each summer movie (not all uploads, just ones with more than 2M Vews). This is the list I had 2 weeks ago

 

GOTG 2 - 65M

King Arthur - 36M

Alien - 28M

Pirates - 46M

Baywatch - 30M

Wonder Woman - 58M

Mummy - 45M

Cars 3 - 40M

TF5 - 32M

DM3 - 52M

Spider-Man - 58M

Apes - 28M

Dunkirk - 28M

Valerian - 16M

 

Transformers 5 looks pretty awful in terms of trailer views to be honest. Might end up being the biggest domestic casualty of the year.

 

32m isn't terrible for views.  However, you also have to factor in that trailer views seem to inflate every year.  I'd assume a drop from a 100m OW, but somewhere between 50-80m OW sounds right for it atm.

  • Like 1
Link to comment
Share on other sites



2 minutes ago, DAJK said:

Two words. Max Steel ;)

 

Yeah, that got factored out as it wasn't a top 50 grosser.  But there were outliers, high trailer view counts are by no means any guarantee for a strong OW or DOM take.  But you can usually use some common sense on when something's an outlier or not (such as Max Steel).

  • Like 1
Link to comment
Share on other sites





22 minutes ago, The Panda said:

 

Yeah I was only measuring YouTube trailer views and DOM Gross/OW.

 

Sometime in the future I may try and do a better test with more movies (top 200), and add in a few extra variables like review, whether it's a franchise movie or not, whether it has Oscar nods or not, etc.

 

This analysis was pretty basic.

 

I am hoping to do something similar for my Quant class in a few months so maybe we can compare 

  • Like 2
Link to comment
Share on other sites



1 hour ago, Jason said:

@The Panda

 

:bravo:

 

I'm wondering if there's something in common with the group of about a dozen films that is well above the overall trendline compared to the rest that track below the trendline (especially those with at least 20M+ views).

 

Im thinking your larger hits and top 10 of the year off the top of my head, but I'd need to double check the spreadsheet.  The actual trend line might actually be a bit more exponential than linear (with Sing being a weird outlier).

  • Like 1
Link to comment
Share on other sites





Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.



  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use and Guidelines. Feel free to read our Privacy Policy as well.