This is a visualization that I created for a recent article in Movies in Data.

It uses a 10 year EWM window wherein a star’s cumulative box office is scored according to their casting precedence in the movie.

The below plot includes all non-animated movies in the dataset (over 500 000 movies). As you’ll see, the biggest movie stars tend to be those who lead a major franchise. This trend begins in 1962 with the release of the first James Bond movie, Dr. No.

actorHorseRaceBoxOfficePowerRaw.mp4

If we remove IP, a different picture emerges.

actorHorseRaceBoxOfficePowerNonIP.mp4

If you’re interested in learning more, check out the article here or the original code here.