Baseball, and especially pitching, is more and more about specialization, and the most specialized case for pitchers are the left-handers who are brought in to face just one batter. The Loogys. The problem with this kind of specialization is that conventional pitching stats are kind of useless for guys like this. Who cares how many earned runs per 9 innings a guy gives up, when you only use him for one batter per game? And how fair is it to measure his walks plus hits per inning pitched, when in most of his appearances he’s either giving up a walk, a hit, or pitching one third of an inning. And even more unfair is that very often these guys are sent out to face one of the toughest outs in the opposing lineup.
So what we need is, a stat that combines whether or not they got the batter out, and how hard that is to do. I call this stat PORK, for pitchers out risk classification. (It’s not a perfect acronym). To start with, we need to figure out how hard it is to get each batter out. I’m simply using each batter’s outs per plate appearance rate. Then, I need to figure out how much it should be worth to a pitcher to get each batter out. To do that, I take the batter’s outs per plate appearance, and divide it by the league average, and take the reciprocal, so guys who make outs less often than league average will be above one, worse batters will be below one, and if someone makes outs at exactly the league average, they’re just worth 1. Through August 29th, getting Andrew McCutchen out once is worth 1.17 on this scale, and getting James Loney out is worth .939. So once we’ve established that, we can sum up a pitcher’s entire season of one batter-faced appearances by taking the sum of each batters out difficulty, and multiplying it by the number of outs the pitcher gets from them. The give the pitcher a bonus for getting a double play off a batter who is tough to get out at all. Then, to create an average I divide that by the sum of the reciprocal of each batter’s out difficulty. I’ll try to explain this better another time.
The point is, looking exclusively at one-batter-faced appearances by lefty relievers, and pitchers who have had at least 10 such appearances this year, Scott Elbert of the Dodgers has been the most effective. He’s had 11 appearances where he faced one batter, and gotten the batter out 9 of those times. And 2 of those 9 were double plays. He’s gotten 5 outs in 4 appearances against batters who are at least 10% better than the league average at not making outs. He got Joey Votto (1.251 out difficulty) once, and 4 outs in 3 appearances against Carlos Gonzalez (1.109). Jon Jay (1.104) is the only “tough out” that Elbert has failed to get this year. All 11 batters he’s faced in loogy appearances have been harder than average to get out, and his total work in those situations give him a PORK of 1.17. Matt Reynolds is next behind him with a 1.09 PORK, and he’s only failed to get his one batter out once in 12 chances. But, he scores lower because he only got one “tough out” (Prince Fielder – 1.123), and the batter he failed to get out should have been on the easy side – Ike Davis, at .960. The group of batters he faced just weren’t as tough as what Elbert went up against.
Javier Lopez has the most 1BF appearances this year among lefties, with 28, and he’s got an impressive .97 PORK in those appearances.
Here are the lefties who have had at least 10 appearances where they faced just once batter, and how they rank in PORK.
As you can see, this might not turn out to be Wesley Wright’s specialty for the long term.
The nice thing about this stat, though I realize it’s a bit convoluted, is it works for comparing any pitching category – starters, relievers, 9th inning with a small lead only guys, 8th inning with a small lead only guys, anything really. It’s not that interesting for Starters though, because it just tells you that Verlander and Felix Hernandez and Matt Cain are really good, and we already knew that. I’ll be posting more about this eventually.
Plunk Signal