carlos salgado
Sep 2, 2018 · 2 min read

Hey. Great article, congratulations.

I’m chilean and I want to study Statistics, after I realized that’s what really excites me, more than a Civil Engineering, for example.

For the last time I’ve been analyzing this kind of data and thinking about what site is the better, between IMDB, MetaScore and RottenTomatoes to make an unique order from these all sites, haha. RottenTomatoes never liked, because of what you say, not common ratings and difficult to interpret. Whit the two others, I’ve been trying to make a Movie/TV Show ranking with its own ratings. For one side IMDB considerates the “crowd feel” and I think that’s important because all of us are part of that and can be represented on it, adding that the number of votes is extremely high, with hundred of thousands and millions, which turns IMDB a very reliable source. In the other way, Metascore adds some criteria from a more analytic point of view. I think is really interesting to trying gathering this two populations of data and see what comes out, that excites me, haha, and in free times I play with their data. Well, I’ve done it and pictured the normal distributions (in both, IMDB and Metacritic, results something you can approximate to), then typify, see how much each rating move away from average in terms of standard desviations,and then to averaging the values of IMDB and Metascore to get an unique number than describes how good is the movie in IMDB and Metascore, 50%-50% measured. After that I got an order of the 100 movies I analyze. The thing is that I’m aware that you have to be very precise with numbers and cannot assume that you have some distribution when you don’t, and maybe I’m making mistakes in that sense. I know that I do this in a very rudimentary way, hehe, but it excites me to learn how to use maths to manipulate this kind of data.

What tool can you recommend me in order to get better my process? Have you thought doign something like this, to gather IMDB/Metacritic ratings?

Regards