This headline has a causal statement (I teach causal inference in econ/ finance for a living, which I love doing) on #Zacksnydersjusticeleague (a movie I love). So I wanted to do a 🧵 explaining whether this statement can even be made with the data we have. 🚨 Long thread ahead! https://twitter.com/cnbc/status/1385213892062810114
If you enter into the article, the first thing that you find are these key points. So, apparently the headline comes from the fact that HBO Max added ONLY 2.7M subscribers during Q1 2021 and this quarter coincides when the #SnyderCut was released.
So, there are two separate points that we can discuss:

1- Was the number of new subscribers into HBO Max low during q1 2020?
2- Can we link the new subscriber performance of HBO Max exclusively to #Zacksnydersjusticeleague ?
Regarding (1), you can compare these numbers with other quarters, with previous years, with other companies. Or as many financial economists do, you can let the market tell you whether these numbers were good or not.
If these numbers implied good news for HBO Max, you should expect the AT&T (which owns HBO Max) stock price to go up. Otherwise it should stay the same or if these numbers are bad, the stock price should decline.
So, what happened to the AT&T stock after the release of these numbers? And how does it compare to Netflix recent release of numbers for Q1? Well the picture is pretty clear. AT&T stock is up by almost 4% today. Netflix after its subscriber growth release tanked almost 8%.
Could it be that there were other news by AT&T that are not HBO Max related that are driving the stock price up? Could be but it’s unlikely. When you look at financial news headlines covering the news, they always signal HBO Max as being the driving force behind the increase.
So, whichever subscriptions HBO Max added, markets seem to think that Q1 numbers were good for the future profits of the company. So, overall these appear to be good numbers. This takes us to (2) and whether we can confidently link these numbers exclusively to the #SnyderCut
This is where things get muddy with the data we have, which is subscriber numbers for each quarter. Can we confidently say that the #SnyderCut drove these numbers instead of Godzilla vs Kong or The Little Things or Judas and the Black Messiah? Probably not.
There might be even subscribers that signed up because all of these movies and not a single one. There might be even #ZackSnydersJusticeLeague fans that signed up in May 2020 when this was announced.
Even for the data for Q4 2020 you cannot even confidently say that it was only about WW1984. Many people might have signed for the same day release of movies for 2021. There might be even DC fans that signed in Dec 2020 because of the possibility of WW1984+ #SnyderCut in March.
Overall my point is that with the publicly available data it is almost impossible to make a causal claim that any particular movie/TV show is alone driving subscriber behavior at any point in time. This is not to say that HBO Max with their own data cannot find this out.
I am sure that their data analytics team is using their private data to try to find out the real demand for each movie/TV project/universe in a way that it is basically impossible to do for us (or any news outlet for that matter).
Overall, HBO Max subscriber numbers in Q1 appear to be solid but it is difficult to exclusively linked them to any particular movie. So do not fall for clickbait titles on either direction.
Final comment on #RestoreTheSnyderVerse: Where the #SnyderCut really excelled and what I think is really promising is in its engagement. With little marketing money and being an “extended version” of an old movie it drove higher numbers than top franchises/movies!
After this long thread on clarifying numbers and clickbait titles, it is a good time for me to go and watch a movie I love about coming together, about not being alone. Time for a #ZackSnydersJusticeLeague rewatch!
You can follow @ClarkJoeDCEU.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: