Box Office and Math

For a while now, I have been getting more and more frustrated with trying to figure out how much a movie will gross on opening weekend. It seems to me, that most people doing predictions are merely making educated guesses. Some are better than others, usually comparing a movie that is coming out with movies that have similar genres and release dates. While this comparison method is better than just guessing, and can give one a general idea of where the movie will end up, surprises at the box office are all too common.

So I thought to myself, that there has to be some other way. A way to take all this information that is available on movies and put it to good use to predict how movies will open. This led me to try and build a mathematical model that incorporates real world variables that I believe are correlated with how much a movie will gross on opening weekend.

Currently, I only have 50 observations which statistically speaking isn't that many. I certainly wouldn't trust a poll with only 50 observations. However, with all the variables put together, I have been able to identify some general patterns and have been able to make some sense out of the limited data that I do have.

Out of the seven variables that I currently have, one of the least useful seems to be the rotten tomatoes score that a movie gets. I think this probably has more to do with the way rotten tomatoes aggregates and scores reviews from critics, then the actual effect that reviews have on OW. Over time with more observations, the rotten tomatoes score might become more significant with regards to opening weekend gross. As of right now though, I am not too hopeful.

I certainly don't expect this model to be perfect, some movies will always remain unpredictable. Next week, I will start using this model to predict movies and I confident that I am going do better than before and be surprised a lot less.