In an earlier THT Live post, I talked a little about heat maps and how they can help us compare players more easily without getting our hands dirty. I thought I would share another visual that most will find less valuable in comparing players, and more for the fun of visuals (at least it was for me).
Chernoff faces were created in the 1970s by the man they are named for in order to display an array of multivariable data in a single diagram. In R, the default amount is to use 15 variables. Each variable will dictate the height, width or style of the eyes, nose or hair of the face. In this case, I made faces for the top 29 hitters (by WAR) in 2010. I’ve listed below which variables are showing up for each characteristic of the face.
Size of the face—I used wOBA (height of the face), wRC (width of the face) and WAR (general shape of the face) to dictate the size of the face.
Shape of the mouth—Power makes people smile, ask Jose Bautista. I used slugging (height), isolated slugging (width) and total home runs (curve of the smile).
Shape of the eyes—Batters who walk a ton need good eyes. Therefore, walk rates (height) and on-base percentage (width) are represented here.
Shape of the hair—I couldn’t think of anything here really, but hits (height) and RBI (width) are here. I also made the “style” of the hair set to stolen bases.
Shape of the nose—A general idea about high batting average on balls in play is that they are unsustainable (not in all cases, but let’s go with this notion for a second). Thus, if a hitter had a high BABIP, he is lying about his true talent. Higher BABIP, bigger nose (think Pinocchio).
Shape of the ears—Psst…players with high strike out rates have big ears because they love to hear the sound of their bat whiffing.
Seriously, look at that big smirk by Bautista. Or those small eyes from Carlos Gonzalez or Carl Crawford. Also notice those really small heads from those who are more well known for their defensive abilities (e.g., Chase Headley, Stephen Drew, Daric Barton). I could go on and on with these.
If you see any other humorous expressions from these players, I’d love to hear them. Or if you have suggestions on changing the variables to better portray ball players based on seasonal statistics.
If you have more time, or are interested in other examples of Chernoff faces, The New York Times did an article about Steve Wang, a professor at Swarthmore College, who used the concept to portray baseball managers. It’s one of my inspirations for trying the tool out on hitters.