Will “You” Really Like This Movie?

If you reviewed this week’s Objective Top Twenty, you might have noticed something other than five additional movies on the list. You might have noticed that, other than Hidden Figures holding onto the number one spot on the list, all of the rankings had changed.

If you reviewed this week’s Objective Top Twenty, you might have noticed something other than five additional movies on the list. You might have noticed that, other than Hidden Figures holding onto the number one spot on the list, all of the rankings had changed.

A few month’s back I mentioned that I was developing a new objective database to project “really like” movies that are not influenced at all by my taste in movies. This week’s Objective Top Twenty reflects the early fruits of that labor.

The plan is to build a very robust database of all of the movies from the last twenty five years that finished in the top 150 in box office sales for each year . I have 1992 through 1995 completed which gives me enough movies to get started with.

The key change in the “really like” formula is that my algorithm measures the probability that users of the IMDB database will rate a particular movie as a 7 out of 10 or higher, which is my definition of a “really like” movie. The key components of the formula are IMDB Average Rating, Rotten Tomatoes Rating, CinemaScore Grade, and the number of  Academy Award wins and nominations for the major categories and for the minor categories.

In future posts, I’ll flesh out my logic for all of these factors. But, the key factor is the capability to measure on IMDB the percentage of IMDB voters who have rated a particular movie as a 7 or higher. When you aggregate all of the movies with a particular IMDB average rating you get results that look like this sample:

Avg. Rating % Rating 7+
                8.5 92.8%
                8.0 88.8%
                7.5 81.4%
                7.0 69.2%
                6.5 54.7%
                6.0 41.5%
                5.5 28.7%

Note that, just because a movie has an average rating of 7.0, doesn’t mean that every movie with a 7.0 average rating is a “really like” movie.  Only 69.2% of the votes cast for the movies with a 7.0 rating were ratings of 7 or higher. Conversely, every movie with an average rating of 6.0 isn’t always a “don’t really like” movie since 41.5% of the voters handed out 7’s or higher. It does mean, though, that the probability of a 7.0 average rated movie is more likely to be a “really like” movie than one with a 6.0 rating.

These changes represent a start down a path towards a movie pre-screening tool that is more useful to the followers of this blog. It is a work in progress that will only get better as more years are added to the database. But, we have a better answer now to the question, “Will you ‘really like’ this movie?”

***

If you’re going to the movies this weekend, chances are that you’re going to see Blade Runner 2049. The early indications are that it is going to live up to the hype. You might also check out The Florida Project, an under the radar movie that is getting some apparently well-deserved buzz.

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 Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s