I made it through one day of calculus in college before transferring out of the class. To fulfill my two required math credits, I took logic (If A, then B, etc.) and statistics instead. My stats professor was a skeptical old Italian who forced us to question every set of data we ran into while simultaneously trying to teach us regression models and standard deviations.
"Who's paying for the study?" he wanted us to ask ourselves. "What data are omitted? What isn't this data set telling us?"
I find that mindset helpful when looking at baseball statistics as well. While watching the Gold Gloves Tuesday night, where Andrew McCutchen won his first, I was reminded of this excellent column by Will Leitch about the change in the way teams use statistics evaluate their players and the reluctance of broadcasters to keep up. Stats have changed over the years. New ones are introduced, old ones marginalized, and none of them tell the whole story. Should it have been McCutchen or Michael Bourn? Some believe it is clear-cut, others think it's a closer race. We see a similar issue in the AL MVP race. Mike Trout had one of the better seasons in baseball history, and his excellent base-running and defense increase his value on the field. But Miguel Cabrera won the triple crown, others counter; how could he not win? It's RBIs vs. WAR, batting average vs. stolen base success rate.
Forget, for a minute, the fact that offensive numbers inevitably factor into Gold Glove voting. Baseball analysis has progressed past the now-worthless errors and fielding percentage. The folks over at Baseball Info Solutions track every batted ball and record its trajectory, velocity and where it goes on the field of play. Two metrics, Defensive Runs Saved and Ultimate Zone Rating, attempt to quantify how many runs above or below average a player allows. Neither of these tell the whole story either, but they represent an attempt to further our understanding of quantifying defensive performance.
Felix Hernandez's 2010 season, when he won the Cy Young despite going 13-12, displayed a realization that traditional stats aren't always the best tools with which to evaluate players. I'm interested to see the progression of fielding stats -- and the perception of those stats -- as we learn more and more about how to quantify and evaluate defense.
First Published: March 24, 2016, 5:19 p.m.
Updated: December 4, 3120, 3:10 a.m.