May 15, 2013

Pebble Hunting

The Starts That Defied FIP

by Sam Miller

On Monday, against the Royals, the Angels got blown out. Wasn’t even close. A walloping, beat at every facet of the game, no drama whatsoever. Just check out the pitching line:

• 13 strikeouts, no walks, no homers.

Meanwhile, the Royals hitters against the Angels’ pitchers:

• 19 hits, 11 runs, all earned.

Do you think I telegraphed the twist too obviously? Did you spot it a mile away? Have you already figured out that both of those pitching lines are the Angels’ pitchers?

This was interesting enough for a tweet, but probably not an article, until I started playing around with single-game FIPs and realized that this game, the one we just saw on Monday, was something like historic. It is likely the most extreme FIP/ERA discrepancy in one game in history.

It is, at least, without a doubt, the biggest FIP/ERA discrepancy of the past year. That was easy enough to calculate. The Angels’ FIP* was 0.31. Their ERA was 11.00. The 10.69-run difference tops that of the Giants’ pitchers (nine Ks, three walks, no homers in a 12-ER outing in April) and easily crushes that of the Phillies (4.75 FIP, 13 ERA on April 5th) and everybody else.

To go beyond that is complicated, because first we have to decide what we mean by “biggest” discrepancy. The Orioles allowed 30 runs to Texas in 2007. The staff’s FIP that day was only 11 or so. That 19-run discrepancy is bigger, to be sure, but not nearly as misleading.

Trying to use a proportion as our scale is similarly difficult, but at the other extreme. The Angels’ FIP was 1/34th as big as their ERA. It was far, far more misleading than the Phillies’ ~3x line in April, or the Giants’ ~5x differential. But what about the Mets’ loss in 1984, when the staff allowed just two runs—one earned—but had a negative FIP, thanks to 16 Ks (and no walks or home runs) in just eight innings. That differential is less than three runs, no match for the Angels', but infinite in proportion.

Without an easy definition, we’re left unable to simply run a search, do the calculations, and sort for a winner. So this’ll take some feel, and some different types of searches, to see what contends.

First, are there any games in which a team struck out more batters than the Angels did, walked nobody, allowed no home runs, and allowed as many runs? No.

Are there any games in which a team struck out nearly as many batters (10+) as the Angels did, walked nobody, allowed no home runs, and allowed nearly as many runs (10+)? No.

Are there any games in which a team struck out lots of batters (8+), walked nobody, allowed no home runs, and allowed lots of runs (8+)? Yes! Ten of those, six coming in the past decade, but in no case did the team strike out more batters OR allow more runs, so we know right off the bat that none can top the Angels. The closest: an eight-K, 11-run performance by the Nationals (Edwin Jackson starting) last year, or a 12-K, eight-run performance by the Rockies (Jeff Francis starting) last September.

Are there any games in which a team struck out more batters than the Angels, and allowed more runs than the Angels, and didn’t allow a home run, but walked a batter? No.

Walked two batters? No.

Walked three batters? Yes, but struck out exactly as many batters and allowed exactly as many runs, so we know that can’t top the Angels.

Are there any games in which a team struck out more batters than the Angels, and allowed more runs than the Angels, and didn’t walk a batter, but allowed a home run, or multiple home runs? No.

So we’re to the point where the only line that can reasonably compare to the Angels is going to be one in which a team allows more runs, strikes out more batters, and allows some combination of home runs and walks; or allows fewer runs but strikes out way more battters, with some combination of walks and home runs. These are the games that match those descriptions:

2011: Cleveland allows 15 runs, strikes out 14, walks two, allows two home runs.
RA: 15.00 FIP: 3.98

2011: Atlanta allows six runs, strikes out 18, walks nobody, allows one home run.
RA: 6.00 FIP: 0.64

2010: Pittsburgh allows 14 runs (11 earned), strikes out 15, walks three, allows two HRs.
RA: 14.00 FIP: 4.09

2008: San Francisco allows five runs (three earned), strikes out 15, walks nobody, no HRs.
RA: 5.00 FIP: -0.13

1996: San Francisco allow eight runs, strike out 16, walk nobody (one HBP), allow two home runs.
RA: 8.00 FIP: 2.86

That last one is more interesting if you imagine doing this exercise and choosing to exclude HBPs, and perhaps using xFIP. Of the entire lot, there is one compelling argument against the Angels: that 2008 game involving the Giants, started by Tim Lincecum. That’s the only game in my searching that yields a lower FIP and any sort of offensive success. I wouldn’t hate the argument in favor of that game.

But I’ll stick with Joe Blanton, Michael Roth, and Robert Coello. One reason I prefer to stick with them is that it’s the perfect group to set this record. Blanton, as noted in this Transaction Analysis after his trade to the Dodgers last summer, totally exposes K:BB as a stat (unlike, say, wins) that can be pitched to, at the expense of overall effectiveness. Roth is an 80-makeup, 30-stuff prospect who will probably never have the stuff to be a good big leaguer but who is absolutely not going to let walks be the thing that does him in. Coello, well, I’ve never heard of Coello. I have now. But he’s Robert Coello, you know? Good for Coello, for being part of this game. Good for Coello, for being just absolutely patronized right now by some jerk writer who knows nothing about him or baseball at all.

I watched this game in person, and it felt like Blanton was getting hit pretty hard. I then went home and watched every hit, and it felt like Blanton was getting bled and blooped to death, with some hard-hit balls in the middle. Rating the 12 hits he allows on a totally subjective 1-to-10 hardness scale, I’d go like this: 3, 7, 1, 4, 8, 6, 1, 8, 4, 4, 3, 9. Two of the hardest-hit balls were misplayed at the wall in right field; there was a bunt single; a ball that hit off him; some grounders; a broken bat; and a couple line drives and a gapper. Furthermore, Blanton got 17 swinging strikes, which is a lot of swinging strikes or, adjusted for Joe Blanton, a ton of swinging strikes. There have been four other starts this year in which a pitcher allowed 12 hits or more, and those four pitchers got three, six, seven, and (Blanton again) seven swinging strikes. You could easily, easily imagine Blanton getting out of this game unscathed, and you can easily, easily see how Joe Blanton is seven strikeouts, no walks, and no home runs away from getting blown out.

I guess I’m saying there’s nothing all that significant about this game. But there’s also nothing all that significant about most perfect games—competent pitcher has a great night, with a ton of luck. And yet, there is something significant about the cluster of them we’ve seen. And there is something significant about this game happening now, and to the starting pitcher it happened to. It’s simultaneously inevitable and a wild fluke.

*For convenience, I used a very simple, standardized FIP formula across all seasons. 

Sam Miller is an author of Baseball Prospectus. Follow @SamMillerBB
Click here to see Sam's other articles. You can contact Sam by clicking here