# ERA, FIP and Kershaw

Joe Posnanski | Sep 3, 2014 |

Sure, I’ll admit it: I don’t know exactly what to do with FIP – the Fielding Independent Pitching numbers that analyze a pitcher based entirely on strikeouts, walks and home runs allowed. On the one hand, I believe the Voros McCracken discovery that most of a pitcher’s value is tied up in those three things, that a pitcher has much more control over strikeouts, walks and home runs than anything else.

Of course a pitcher doesn’t have complete control of ANYTHING, and no matter who the pitcher may be Ryan Howard is going to strike out a lot, Jose Bautista is going to walk a lot, and Giancarlo Stanton will hit many home runs. But these are the three areas where I think everyone can agree a pitcher has significant say in the matter.

It’s considerably less clear how much control a pitcher has of a ball put in play. The fate of balls in play seem to be more in the realm of defense and luck and the quirks of a particular ballpark. Then again, it’s hard and maybe incorrect to make the leap that a pitcher has NO say in those balls in play. People will argue about this all the time.

The point is – I like FIP a lot. FIP is the Tom Tango and Clay Dreslough invention that essentially crunches strikeouts, walks and home runs into a number that is aligned with ERA. The basic formula is pretty easy:

((13 x home runs) + (3 x walks + HBP) – (2 x strikeouts)) / innings pitched.

That gives you a raw number where 0 is pretty darned good. A pitcher would like his raw FIP to be even less than zero.

Three pitchers here:

**Corey Kluber, Cleveland** -- 14 homers, 46 walks, 6 HBP, 215 Ks, 195.3 innings. Raw FIP: -0.47

**Rick Porcello, Detroit** -- 15 homers, 33 walks, 3 HBP, 109 Ks, 180 innings. Raw FIP: +0.47

**Shelby Miller, St. Louis** -- 20 homers, 68 walks, 2 HBP, 101 strikeouts, 152 2/3 innings. Raw FIP: +1.76

OK, so you can see based on this – Kluber has a very low FIP, Porcello has a pretty decent FIP, Miller’s FIP is pretty much disastrous. Now, what Tango and company did to put the number in context and make it seem more familiar and friendly – they added a constant that is tied directly to the league’s run scoring environment. That way the numbers should correspond, more or less, with ERAs.

You can see it how FIPs look like ERAs when you look at the three above pitchers:

Kluber: 2.58 ERA; 2.66 FIP. Porcello: 3.10 ERA; 3.60 FIP. Miller: 4.19 ERA; 4.88 FIP.

So the numbers more or less match-up. All three pitchers have FIPs higher than their ERAs, but Kluber’s is barely higher, Porcello’s is a half run higher, Miller’s is almost three quarters of a run higher.

As I say, I really like using FIP. But I can’t quite cut the cord with ERA – can’t quite make the leap that how many runs a pitcher allows (or, to me more precise, how many runs a team allows with a pitcher on the mound) should not be part of the equation. In a way, I suspect this is probably a lack of imagination on my part, tied up in my childhood. But, you can’t relive your childhood. I grew up with ERA as THE thing, and it’s hardly the only childhood thing I can’t let go of. Point is, I tend to look at both FIP and ERA when looking at a pitcher.

Which leads me to this: Clayton Kershaw is having a crazy, historic season.

OK, you probably already knew that. But this morning, I was looking at his ERA (1.70) and his FIP (1.89) and I thought: Wow, that’s probably pretty unusual for a pitcher to have his ERA and his FIP less than 2. I wonder who the last person to do that was. My guess was Kershaw himself last year.

But … no. Kershaw last year had a 1.83 ERA but his FIP was 2.39.

OK, so I thought a bit more. Well, obviously Pedro Martinez did it.

Nope. The two years Pedro’s ERA was less than 2.00 (1997 and 2000) his FIP was a little bit above 2.00. And the crazy years when his FIP was less than 2.00 (1999* and 2001 in limited innings) his ERA was a little bit above 2.00. Now, this was getting interesting.

**In 1999, as you probably remember, Pedro pitched 213.3 innings, struck out 313, walked 37 and allowed nine home runs. That was an utterly absurd raw FIP of minus-1.74 – the lowest raw FIP in baseball history by a long, long shot.*

Here are the Top 5 in Raw FIP:

1. Pedro Martinez, 1999, -1.74 2. Clayton Kershaw, 2014, -1.23 3. Dwight Gooden, 1984, -1.08 4. Matt Harvey, 2013, -1.04 5. Randy Johnson, 1995, -1.02

*Pedro’s ERA that year was 2.07, which was utterly absurd in the 1999 American League.*

If it wasn’t Pedro, it had to be Greg Maddux, right? Nope. Maddux never had a sub-2.00 FIP. Roger Clemens? Nope. Clemens never had a sub-2.00 FIP. Randy Johnson? Nope.

OK, well, then it had to be Dwight Gooden in the mid-1980s when he was basically unhittable. Except – no. Gooden is fascinating. His legendary season is 1985, when he went 24-4 with a 1.53 ERA and led the league in strikeouts. But, crazily, his best FIP season was actually 1984, when he struck out 276 in 218 innings and allowed just seven home runs all year. His total FIP that year, 1.69, is the second best since Deadball, only behind Pedro’s insane 1999 season.

So now what? Ron Guidry in 1978? Nope. His FIP was 2.19. Steve Carlton in 1972? Nope, just missed, his FIP was 2.01. Gaylord Perry in 1972? Nope. So how far back do you have to go anyway?

Well, you have to go back one more year – Tom Seaver in 1971. He had a 1.76 ERA – and thanks to a league leading 289 Ks, 61 walks and 18 homers allowed, he had a 1.93 FIP.

In total, Kershaw is on pace to become just the fifth pitcher since Deadball to have a sub-2.00 ERA and FIP. The previous four are all-time seasons:

1946: Hal Newhouser, 1.94 ERA; 1.97 FIP 1963: Sandy Koufax, 1.88 ERA, 1.85 FIP 1968: Bob Gibson, 1.12 ERA, 1.77 FIP 1971: Tom Seaver, 1.76 ERA, 1.93 FIP. 2014: Clayton Kershaw, 1.70 ERA, 1.89 FIP.

Amazing stuff. Kershaw should become the first pitcher in more than 40 years to tilt ERA and FIP, only the fifth ever, a year up there in its own way with Gibson’s 1968 season. There is no shortage of ways to show just how awesome Clayton Kershaw is these days … but I like this one. Kershaw is dominant in old stats and in new ones. That could be what they mean by timeless.