Recently I mentioned to someone that I was a movie blogger. Naturally they assumed I wrote movie reviews. It did get me thinking, though, “what is my blog really about?”
Looking back at my previous 92 posts, it’s hard to discern a consistent theme. I confess that it has been hard to come up with a fresh idea every single week. The result has been a hodgepodge of articles that intersect movies and data, but lack a unifying core. That is…until now.
It occurs to me that, while I’m not in the movie reviewing business, I am in the movie previewing business. I use statistical analysis to preview what movies I might “really like”. It also occurs to me that I created my algorithm for my benefit, not yours. I write this blog, though, for your benefit.
With all of that in mind, I’ve decided to reorient this blog to a discussion of movies you might “really like”, using my statistical analysis as the foundation of the discussion. My algorithm has two parts. The first produces a “really like” probability based on data from websites like Netflix, Movielens, and Critcker that are oriented to movies that I, personally, will “really like”.
The second part of the equation is based on general data that has nothing to do with my personal taste in movies. IMDB and Rotten Tomatoes produce ratings based on the input of all of their website participants. Oscar award performance has nothing to do with me. I’m not a member of the academy. For now, these are the factors that go into my “really like” probability based on general information. It’s this “really like” probability that might be most useful to you, the followers of this blog.
On Monday I added a new list to this site. The Top Ten 2017 Movies Based on Objective Criteria uses this second half of my algorithm to suggest movies that you might “really like”. I intend to update this list every Monday after the initial data from the previous weekend’s new releases comes in. This Friday, for example, Wonder Woman goes into wide release. Some critics are calling it the “best superhero movie of all time”. It will be interesting to look at the data feedback on Monday to see if it’s actually trending that way.
I’m also exploring the addition of other general data to the criteria. For example is there statistical significance to when a movie is released. I’m in the data gathering stage of that study. I’m also planning on adding in future months Top Ten lists for years prior to 2017.
I will also continue to update on Wednesday’s my Watch List for the week. While it is based on movies I should “really like”, you might find some movies there that pique your interest.
As for this blog, I plan to orient each week’s post around one or two of the movies on my lists and offer up some ideas as to why it might be a movie that you’ll “really like”. For now I would encourage you to check back on Monday to see if the hyperbolic buzz surrounding Wonder Woman is supported by strong enough numbers to move it into 2017’s “really like” Top Ten. Then, return again on Thursday to see what movies that you might “really like” have caught my eye.