16 March 2014

The Next Frontier In Baseball Analysis

Okay, so that's settled. Baseball front offices have completely bought into the new analysis that values on-base percentage, slugging percentage, BABIP, stolen-base efficiency, defensive efficiency and defense-independent and park-independent pitching statistics rather than batting average, homers and RBIs, raw steals, fielding percentage and pitching wins. The sports media and ordinary fans are coming around too.

In the marketing field, they say that a corporate tagline begins to become effective just about the time the company gets sick of it. In baseball, the same thing is happening. No sooner has all of baseballdom finally acknowledged that the number crunchers were on to something (save for a handful of writers like Murray Chass and Dan Shaughnessy who have carved out a niche clinging to the Stone Age) than the seamheads have moved on.

For the past seven seasons, MLB ballparks have been equipped with cameras capturing the path, break, speed, strike zone position and result (ball, strike, hit, out, etc.) of every single Major League pitch (Pitch/FX). The information that can be gleaned from the data beggars the imagination. The percentage of pitches thrown by any given pitcher that are balls or strikes, swung at or taken, the mix of pitches, etc. are now public knowledge. Anyone with too much time on their hands can mine the data for percentage of pitches a pitcher threw outside the strike zone that were offered at by the batter and what percentage of in-zone pitches were taken for strikes. Anyone with 25 spare hours a day can compare how a pitcher approaches righties as opposed to lefties, day games versus night games, early innings versus late innings, etc. (Ditto for batters.)

Below is just one of the many charts, graphs and tables that comprise Pitch/FX's treasure trove of material (Thank you Brooks Baseball and Baseball Prospectus.) It documents Craig Kimbrel's pitch selection (one change-up his entire career!) and the average velocity and movement of each pitch type. As you can see from the headings, this is a mere tip of the iceberg.



Pitch/FX also allows us to follow the results of these pitches, as seen below in this swing rate graphic. Note how Kimbrel induces 60% of batters to take their cuts at pitches down the middle but above the strike zone. (The percentage of pitches thrown outside the zone that batters swing at is the O-swing rate. Good pitchers tend to have high O-swing rate. They are fooling people. Batters with high O-rates may be good hitters but are not very selective.)


Batters also take for strikes a third of the pitches he fires right down the middle. The percentage of pitches in the zone swung at is called the Z-swing rate. Good pitchers have relatively low Z-swing rates. Batter with low Z-rates tend to be selective hitters who walk and strike out a lot.


This information is incredibly useful for evaluating pitchers and hitters because it reveals a heretofore hidden part of the game. It is likely that Pitch/FX could have showcased and explained the special genius of Greg Maddux, whose shockingly ordinary repertoire yielded league-leading strikeout totals, probably because of a low Z-rate swing % (i.e., lots of called strikes) and a high O-rate swing % (lots of weak contact.) You can just imagine the on-field adjustments that can be wrangled out of this vast ocean insight.

The same information can disassemble a batter just as quickly. Teams are pouncing on the goldmine of data Pitch/FX is uncovering. And starting this season, there will be even more gold to mine.

Teams are now placing cameras afield to capture the movements of defenders and compare them with the most efficient path towards the ball. These cameras will monitor jump, speed and path of the defender and the speed, path and height of the ball when (if) caught. It is the next frontier and it promises quantum leaps in accuracy of defensive statistics. It will tell us why, for example, Jacoby Ellsbury's blazing speed doesn't translate to more fielded balls than average. And it will reveal the special genius of Derek Jeter, who will be getting out just in time.

All this will increase the premium of smart front offices and give low-income teams like Tampa opportunities to win with lower-cost talent. It could also reduce parity as front-office prowess, perhaps particularly in the number-crunching departments, rises in importance compared to on-field ability.

In any case, it's a bold new world and more than ever, those slow to adapt will feel the wrath of the competitive marketplace, even as they earn hosannas from dinosaurs like Chass and Shaughnessy.


No comments: