Monday, August 26, 2013
Should reaching on errors count as hits?
By David Schoenfield
A few weeks ago, Jared Cross, who helped develop the Steamer projection system, wrote an ESPN Insider piece comparing the little advantages that Mike Trout held over Miguel Cabrera -- things like reaching base on errors, not grounding into double plays in double-play situations, hitting sacrifice flies or groundball outs that at least advance a runner.
At the time (the piece was published on Aug. 6), these little things added up to about 13 to 14 extra runs created for Trout compared to Cabrera -- or little more than a win's worth of runs.
I read through the comments and, frankly, they were a little depressing, with one common argument going something like, "This just shows what you idiot sabermetricians will do to try and prove Trout is better!"
Here's the thing about that complaint: Aren't the "little things" actually an old-school argument about players? You know, "He does the little things to help his team." I remember years ago I had to do a presentation to a group of people who worked on ESPN's baseball coverage, including some of the analysts then working for the network. I was discussing the merits of OPS as a better way to measure a player's offensive value than just batting average or RBIs; one of the former players suggested the stat was biased against the players who did the little things.
Well, you know, that player was actually kind of right; the impact isn't significant as that player probably wanted to believe, but the little things -- as Jared quantified -- can add up to create some value.
Trout's biggest advantage came via reaching base on error, worth 7.7 runs compared to Cabrera. Now, when a player reaches base on an error, he doesn't get credit for it. In fact, it counts against him -- it's counted as an out on his ledger. As Joe Posnanski wrote last week, "It's one of the dumbest statistical tricks in all of sports, maybe the dumbest ... If you hit the ball and reach base it should absolutely NOT be counted as an out. It’s not an out. No out was recorded. IT IS NOT AN OUT. Sorry, I am going off on a rant here."
Joe pointed out that, at the time, if you counted Trout's ROE as hits his batting average would be .350 and his on-base percentage .444. Right now, his average would be .343 based on the 10 ROE that Baseball-Reference credits him with (ESPN credits him with nine, more on that in a bit).
Even if you don't want to count a ROE as a hit, there is some logic to counting it towards a player's on-base percentage, in part because fast players tend to reach on errors more often than slow players, or so goes the theory. According to the ESPN Stats & Info database, the players with the most ROE this season: Andrelton Simmons (10); Trout and Norichika Aoki (9); Robinson Cano, Jay Bruce and Elvis Andrus (8); Starlin Castro, Jon Jay, Jean Segura, Eric Hosmer, Chris Denorfia and Alexei Ramirez (7). All with the exception of Bruce run pretty well.
Since 2009, the players with the most ROEs are Andrus, Marlon Bryd, Michael Cuddyer, Ramirez, Martin Prado, Brandon Phillips, Ryan Braun, Cano, Ichiro Suzuki, Michael Young, Hunter Pence and Adam Jones. Again, other than Cuddyer, I'd say all of these guys have at least average to above-average speed.
OK, back to Trout. I thought I'd check his ROEs this season and see if his speed actually came into play -- did the fielder rush his throw, for example?
April 28 -- Fielding error on Mariners shortstop Robert Andino. Slow roller that Andino had to charge. He actually looked like he was thinking of going to third to get the runner there but didn't come up with the ball. He had no shot to get Trout at first, however, and not sure if he would have gotten the runner (Andrew Romine) at third.
April 30 -- Throwing error by A's shortstop Adam Rosales. Chopper that Rosales had to charge and hurry his throw. A good throw gets him but Trout's speed no doubt forced a quick throw.
May 9 -- Throwing error by Astros third baseman Matt Dominguez. Routine grounder. Just threw it away.
June 17 -- Throwing error by Mariners second baseman Nick Franklin. Grounder way Franklin's left in shallow right, off-balance throw was in the dirt. Maybe has time to set himself with a slower runner.
June 19 -- Fielding error by Franklin. Franklin again ranges to his left but the ball goes off the heel of his glove. The Angels' analyst brings up the whole "with Trout running you're thinking of his speed" thing.
June 25 -- Throwing error by Tigers third baseman Miguel Cabrera. Cabrera thought about tagging the runner moving to third and that moment of hesitation cost him as he then rushed his throw to first and threw it in the dirt.
June 26 -- Fielding error by Cabrera. Boots grounder right to him.
June 29 -- Throwing error by Astros shortstop Jake Elmore. Chopper to Elmore's right, had to hurry the throw. Astros announcers bring up Trout's speed as a possible cause.
July 2 -- Fielding error by Cardinals second baseman Matt Carpenter. With a runner on first, grounder up the middle that Carpenter bobbled and then threw to first, too late to get Trout. This is the play that Baseball-Reference classifies as a ROE but ESPN doesn't (if he fields the ball cleanly he probably gets the force play at second).
Aug. 24 -- Fielding error by Mariners right fielder Endy Chavez. Had to run a long way but dropped a fly ball.
So that's nine (or 10) ROEs. I'd say that Trout's speed came into play on six of them, maybe seven if you want to include the Carpenter play. I'd say two of the plays -- the Andino error and the first Franklin error -- are often counted as hits.
Anyway, at least with Trout, it seems clear in watching the evidence that his speed has factored into his ROEs; it's not just random luck. Maybe we don't get as far as Posnanski suggests and count them as hits, but I'm in agreement that ROEs should at least count towards a player's OBP.
What do you think?