11940 posts out of which 1069 posts were the popular ones.
Analyzing Medium’s posts and building a simple prediction service for “Popular on Medium”
Shitij Goyal

Given your statistics, it seems more than 90% of all articles are not popular. That means a 90% accuracy of your model is not impressive, since you get that by classifying all articles as non-popular. Did you check how many of the popular articles were correctly classified?

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.