The Problem With ERA
Beginning yesterday, we’re trying something new and a little different with the guys over at Brock and Salk – we’re doing some post-swapping. I’ve agreed to write one post per week for their site, and Mike will contribute pieces to USSM on a semi-regular basis. The idea is to give both of our audiences something that they don’t necessarily get right now; I’ll be bringing some statistical analysis to their blog, and Mike will bring some of the behind the scenes stuff to ours. Both perspectives are valid, and we think it will give everyone a more balanced view of the game.
Anyway, my first post at their blog is up now. It’s a pretty basic primer on the flaws of ERA, so it may not be news to you, but I tried to write it in a way that wouldn’t be intimidating to people who don’t like math. We’ll have Mike’s first USSM post up here in the not too distant future.
Hope you guys enjoy the collaboration.
Awesome for us and what a boon for KJR listeners…I’m happy to read what Mike thinks is a good piece from his world for this audience, should be interesting…so long as we don’t have to read anything by Huard I’m sure I’ll really enjoy the swaps. Neat idea.
Dave, when are you finally going to decide you’re overwhelmed with writing gigs and have to turn something down?
Great idea. I don’t always agree with him, but Salk is easily the best thing to happen to sports radio around here for years. He is intelligent, and doesn’t mind admitting when he is wrong or uninformed. And he is damned funny.
Nice piece. I wish it could reach a wider audience (say, on ESPN.com). Unfortunately there are many who maintain that win-loss record is also a solid indicator of a pitcher’s ability. Ugh…
As for the pitcher being the sole party responsible for home runs–tell that to the hurler who had Jose Canseco’s head as part of his outfield defense.
So when do pitchers who are considered “unlucky” due to a high BABIP are just considered bad? Because I know Justin Masterson is considered a very unlucky pitcher with his high BABIP but is he honestly good? I assume .300 is a benchmark for if a pitcher is lucky/unlucky.
You can be both bad and unlucky. If your luck turns around, that doesn’t make you good — just less horrendously bad. That’s what the advanced pitching metrics try do determine: by factoring out (or normalizing) the aspects a pitcher can’t control and only looking at the aspects he can, you hope to find his “true” talent level.
And yes, we generally expect a pitcher’s BABIP should regress to around .300. But it may take quite a while for it to do so when a starter might only see 20 or 30 batters in a typical outing, and not all of them put a ball in play; relievers of course see even fewer. So luck can color your perception of a pitcher for a long time after you think you know what he “really” is.
Interestingly, last year Dave found a small but consistent spread between home and road BABIP for pitchers. There are other exceptions to the “about .300” expectation: knuckleballers tend to have abnormally low BABIP (Wakefield has had his BABIP hit .300 the last couple of seasons, but it was low enough over the previous years for his career average to still be .282; Charlie Hough held his to .258), and certain other “effectively wild” pitchers seem to possess that trait as well (Nolan Ryan had a career BABIP of .275)
I should add wrt Masterson in particular that he shows significant platoon splits, and while he’s been carrying a high BABIP against both lefties and righties, it’s going to hurt him much more with lefties whom he strikes out at less than half the rate he manages against righties (his HR and walk rate are much higher against LH batters as well).
This is reflected in his platoon xFIP (5.01 vs L, 2.21 v R) — it’s probably safe to say that facing RH batters he’s reasonably good but unlucky; against lefties, he’s both unlucky and bad. Unless he develops a changeup or some other pitch he can throw effectively against LH batters, he’s not a particularly effective starter.
(Usual sample size caveats apply, though the split is apparent in his career stats also.)
I think this is a great idea (sharing your writings with each others’ blogs). Keep it up. The only issue I see is I just can’t keep adding more blogs to read/check everyday. It gets overwhelming. Can you drop a line on this blog to let us know when you have something on another site? I read five baseball blogs right now, and that’s just baseball not any other topics.
Keep up the great work……and writings.
The comments about his splits are really interesting, but let me respond to the more general point. One way is to wait until his BABIP is normal (which will eventaully happen after enough time) and enough time has elapsed that his ERA will reflect his real talent level. That takes about 2 or 3, and sometimes even then it doesn’t happen (see Vazquez, Javier).
Another way is to look at his “three true outcomes”–walk, homerun and strikeout rates. In the course of about 100 IP, these three things are typically reflective of the pitcher’s real skills and give you a good idea of his talent level. They’re also pretty stable from one season to the next (when they’re not, they usually predict a breakout or collapse!) There’s a mathematical function that combines these three stats to give you a measure a pitchers skill:
This number is called FIP (Fielding Independent Pitching) and it gives you a number roughly on the scale of ERA, i.e., it tells you how many runs will cross the plate based on these three stats alone. Cool fact: after about 50 IP, a pitcher’s current FIP will better predict his future ERA than his current ERA will.