Tuesday, July 21, 2015

Defensive skill differences aren't important in Major League Baseball

Baseball’s statisticians have long been looking for a way—any way—to figure out what a player is worth on defense. It was nothing less than the holy grail of baseball statistics.

...the nerds went to work, coming up with computerized statistics that measured range and outs created, boiling the entirety of defensive performance down to a single number. Problem solved.

But now there’s another problem. A new trend—one that ironically also emerged from baseball’s quest to maximize defense—is rendering all that hard work moot: the defensive shift.

Many of the newfangled metrics that were built to judge a fielder’s range largely ignored where the player was actually standing when the ball was hit. So, the more teams shift defenders in between at-bats, the less the statistic makes any sense. ...

The same batted-ball data that enables range and defensive skills to be quantified also tells teams where to position their players for each batter. This is making it almost impossible to assign proper credit for a great defensive play: Did a player make a play because he has incredible range, or because prescient scouting had him stationed in that area?

The result is something that no one would have predicted: Eyeball scouting may be more necessary than ever. ...

Inside-Edge, a baseball analytics company that provides data to major-league teams, brings a big-data approach to scouting. Instead of just using a spray chart to calculate a player’s defensive value, their scouts watch every single play from every single team—twice. ...

The major revelation: The quality of a fielder doesn’t matter on most plays. Inside-Edge partner Kenny Kendrena says 24% of plays are almost always hits and 62% are almost always outs. ...

Using the eyeball test to study how far a fielder has ranged to make a play—rather than just looking at where on the field he was—significantly cuts into the number of hits that fielders are given credit for impacting. For example, if Colorado shortstop Troy Tulowitzki retrieves a ball behind second base, zone statistics might look at where he was and reward him for making a good play that saved an out. But the human eye may reveal that Tulowitzki started a few feet from where he ended up and deem his efforts routine. With this methodology, Inside-Edge credits major-league teams for saving 179 hits over what was expected with an excellent play as of Tuesday, while bad defense was blamed for allowing 156 hits that normally would be outs.

But that seems like a drop in the bucket given that position players have had 76,112 total chances this season. Baseball’s best defensive team, the Houston Astros, has saved only 25 hits all season because of the skill of its defenders, according to Inside-Edge.
--Andrew Beaton and Michael Salfino, WSJ, on Big Data negating Big Data