PCA Outlier Hitters
Baseball collects a lot of data. It’s awesome. FanGraphs is a fun place for data. Exit velocity, spin rate, launch angle; these are fun data points. But, the vast majority of data that is floating around in lakes and clouds is generally not as exciting. Take some kind of machinery for example. Right now there are gears whirling, sensors sensing, detectors detecting, you get the point. This type of data is typically referenced in discussions about the internet of things (IoT). Baseball analytics has always benefited from what is learned in industry and in this post, I’ll be investigating whether a common industry technique, a Principal Component Analysis (PCA), can be useful in baseball analytics.