When It Comes To Movie DNA, Do Directors Have It and Has Netflix Mapped It?

Do movie directors have a movie making DNA? Do movie directors make movies with common traits, a movie making DNA, that particular viewers might be drawn to or repelled by? Netflix has made millions of dollars by identifying movies and TV shows that we are predisposed to enjoy.Does Netflix, indirectly or directly, draw you to favorite directors and push you away from directors you just don’t get? These are the questions I’ve been researching the past week.

I can’t wait to see Christopher Nolan’s next movie, Dunkirk, which is due to reach the theaters in 2017. I’ve seen 8 of Nolan’s 9 feature films and have given those movies an average rating of 87.5 out of 100, my highest average rating for any director with at least 8 movies seen. On the other hand, I’m bored to tears by Wes Anderson’s movies. I’ve seen 2 of his 8 movies and I’ve awarded them an average rating of 43.5 out of 100. Not included in the two movies I watched were Rushmore and Grand Budapest Hotel, both of which I tried to watch but couldn’t finish. Each of us has a distinct movie taste that guides our movie selection. It is our own unique movie DNA, if you will.

Do movie directors have a movie making DNA?  Do movie directors make movies with common traits,  a movie making DNA, that particular viewers might be drawn to or repelled by? Netflix has made millions of dollars by identifying movies and TV shows that we are predisposed to enjoy.Does Netflix, indirectly or directly, draw you to favorite directors and push you away from directors you just don’t get? These are the questions I’ve been researching the past week.

I haven’t successfully come up with a broad systematic answer to these questions yet. But, by looking at a couple of directors, one I like and another I don’t, I can begin to develop a hypothesis. The two directors I looked at have a sizeable body of work. The director who I enjoy is Ron Howard. In the last 15 years I’ve seen 9 of the 21 feature films he has directed.. The director I just don’t get is Stanley Kubrick. Everyone praises his genius but I don’t “really like” his movies. Here are my average ratings for these two director’s movies that I’ve seen over the last 15 years compared to the average ratings of all Netflix customers for the same movies. For purposes of apples to apples comparison, Netflix ratings have been converted to a 100 point scale (e.g.  3.8 out of 5 Netflix Rating is 76 on a 100 point scale).

My Avg Rating Netflix Avg Rating My Rating Difference
Ron Howard 77 76 +1
Stanley Kubrick 52 76 -24

My enjoyment of Ron Howard is fairly consistent with everyone’s enjoyment of Ron Howard. He makes movies that appeal to the general audience. This probably suggests that well done mainstream movies are in my movie DNA. On the other hand, there is a clear difference between my taste for Kubrick and everyone else’s taste for Kubrick. He is not mainstream.

So does Netflix recognize the different appeal that these two directors have for me?  Here’s the same table as the one above, except with the Netflix Best Guess average rating for how I’ll rate the movies instead of how I actually rated the movies..

Netflix Best Guess Avg Rating for Me Netflix Avg Rating Netflix Best Guess Difference
Ron Howard 83 76 +7
Stanley Kubrick 72 76 -4

Directionally it is consistent with my ratings. It is more bullish than my ratings for Ron Howard’s movies and less bearish for Kubrick’s movies. Interestingly enough it is most bullish for Ron Howards best movies as displayed below:

Ron Howard’s Movies I’ve Seen
Netflix Best Guess Avg Rating for Me Netflix Avg Rating Netflix Best Guess Difference
Netflix Avg Rating > 76 92 78 +14
Netflix Avg Rating < 76 72 73 -1

Netflix highly recommends Ron Howard’s best movies to me while taking a neutral position toward his middle of the road movies.

If there is such a thing as a director’s movie making DNA and if Netflix is successfully factoring it into the Best Guess Ratings developed for me, then that DNA relationship should exist in the movies I haven’t seen in the last 15 years as well. Here’s a look at the sample for those movies:

Netflix Best Guess Avg Rating for Me Netflix Avg Rating Netflix Best Guess Difference
Ron Howard 67 70 -3
Stanley Kubrick 55 70 -15

Again, the results are consistent with the results for the movies I’ve seen. My additional observation is that the director I like gets a Netflix recommendation boost for the movies that the Netflix universe rates the highest. Conversely, Netflix more aggressively drives me away from the movies rated lowest by the Netflix universe for the director I don’t like.

Without a broader study, I can’t say for sure that there is such a thing as movie DNA specific to a movie director, or that Netflix’ algorithm indirectly recognizes it in their recommendations. But, based on this isolated comparison, it sure looks like there is and Netflix might have it well mapped.

 

 

Author: Mad Movie Man

I love good movies. In my prior life I worked with predictive models. I've combined my love of movies with my prior experience to create a simple Bayesian probability model to help select movies that you will probably "really like".

Leave a comment