Instructional Video
Horror Analytics Overview
The report below shows the current movies that are in the data set. I recommend watching the instructional video before proceeding to filter any of the visuals in order to give you a better understanding of how the movies are broken out. You will be able to find the adjusted rating as well as ratings from other sources within the report. The chart pertaining to peak points is undergoing additional revisions.
The Horror Report
*Mobile view is not enabled for embedded Power Bi reports, view in landscape, or navigate to page three for the list.
Snapshot of All Data as of June 2023
I intend to update the data regularly, so please check back to see what movies have been added—filter by theme or by rating. If you have any: questions, comments, or suggestions for improvement, please send an email, and I’ll respond as soon as I’m able.
Data Journey… So far
“From the Top
I was re-watching horror movies from my collection, as I often do when I thought about how it seems really hard to find a “good” one. Depending on the source, I was getting different ratings and couldn’t nail down which ones to trust. This was the moment that birthed this entire journey. ”
“What am I Doing!
This is still the question I ask myself, why am I dedicating my time to this hobby? Why would anyone even trust my ratings? I can only really answer the first question, and that’s simply “because I wanna.” I like data analytics and problem-solving, and for me, finding accurate ratings of horror movies was a problem. ”
“First Iteration
I knew the problem I wanted to solve, but now it was time to figure out what data points I would use and how to define them. My first iteration was to look at “peak points.” The moments of tension, suspense, and shock using the timestamp of at the top of the moment. ”
“First Test
What I found with just using the timestamp and intervals without any controls, was that I was getting large gaps or high averages because of exposition. To remedy that, I had to eliminate exposition so I can focus on the real meat of the scares. ”
“Second Iteration
My next issue was the varying length in films. Depending on how many peak points and the length, a rating could be inflated, therefore I had to normalize the peak points for each film. ”
“Second Test
Normalizing the peak points by 2 minutes (average time between climatic scenes) was the best “fit” for the calculation. Once more films are collected, this standard might change.”
“Third Iteration
This is a marathon, not a sprint, and in a recent conversation with another data nerd, I realized that I was missing a key driver in the calculation, quality. This is a tough thing to measure, but at the moment I am going through a process of evaluating the quality of the peak points with a low to high scare metric. To ensure movies aren’t just making the grade due to ratings and peak points if the average of those scares is below a certain threshold then the rating can be reduced. ”
“Dashboard Update
After going through the data set and adding quality metrics to the best of my ability (again, this is going to be a bias because what I would determine to be a low-quality scare, might not be the same for others), I wanted to trim it down to what I thought would be relevant. I refreshed the look for the “Top Film,” added a “Quality Scare” metric, and a new visual to compare other source ratings with the one I developed by element.”