Concerning Jim Johnson and Groundball Relievers in General

Despite leading the AL in saves,  Orioles closer Jim Johnson is having a rough year compared to 2012 when he posted a 2.51 ERA and saved 51 games in 54 opportunities. Early in 2013, an enthusiastic Orioles sportswriter named Johnson the best closer in baseball, a statement that doesn’t look quite so good a few months later. As a closer who relies on the groundball, Johnson is something of an odd bird (pun intended). In 2012 his 15.2 K% ranked 130th out of 136 qualified relievers and his Zone-Contact% was 2nd highest. This year his 18.0 K% ranks 111th out of 140 qualified relievers and his Zone-Contact% is 9th highest. While Johnson has struck out a few more hitters, he has also walked slightly more, from 5.6% to 7.1%, and his groundball rate is down. Overall, his fielding-independent numbers are basically the same as last year. Various explanations have been offered for Johnson’s lack of success in 2013 compared to 2012. Bill Castro, the Orioles interim pitching coach (check out his 1979 season) attributes Johnson’s struggles to overthrowing, and a failure to locate down in the zone which has resulted in less early contact outs. I prepared the following chart to check up on these explanations.

	Bottom Third%	MiddleThird%	UpperThird%	2-Seam velo	Z-Contact%	GB%
2012	14.0	14.5	8.3	94.2	92.9	62.3
2013	14.5	14.4	7.7	93.7	90.6	56.2
Career	14.7	14.6	8.3	94.2	90.2	57.5

So Johnson is throwing slightly more pitches in the lower third of the zone, and actually getting more swings and misses on pitches in the zone. The overthrowing statement seems faulty, as Johnson’s velocity on his sinker is actually down. A look at the Pitch f/x data shows that his sinker has flattened out slightly from last year, though the difference is slight overall. The following chart shows what kind of contact batters are making off Johnson compared to last year.

	BABIP	HR/FB	HR%	HR Per Contact
2012	0.251	6.8	1.1	1.4
2013	0.323	12.5	2.1	2.9
Career	0.286	8.0	1.5	2.0

And to go in even more detail the following two charts show BABIP by zone and then the slugging by zone for Johnson.

		BABIP
	Lower Third	Middle Third	Upper Third
2012	0.289	0.296	0.259
2013	0.292	0.486	0.423
Career	0.286	0.298	0.343
		SLG
	Lower Third	Middle Third	Upper Third
2012	0.294	0.283	0.516
2013	0.321	0.561	0.515
Career	0.331	0.382	0.503

So balls put in play against Johnson have been falling for hits more frequently this year and those hits have been more damaging in each third of the strike zone. In particular, the pitches Johnson has thrown over the middle have been getting hammered. Last year, the results on those pitches were quite tame. Granted, this is a pretty small sample size of balls in play, and nowhere near the point where BABIP is expected to stabilize, but it goes to show that Johnson has not fared nearly as well when hitters are making contact in 2013. But, this is not an uncommon issue for high-contact, groundball pitchers. David Robertson can suffer through a .335 BABIP in 2012 and still post a 2.67 ERA on the strength of a 32.7 K%. Pitchers like Johnson who cannot strike out hitters regularly are subject to variance on batted balls. Take a look at most groundball, contact-type pitchers, and you’ll find years where BABIP and ERA go through the roof. With the 60-70 inning seasons relievers work, the results can get skewed very badly. To get a sense for where Johnson stands relative to other groundball relievers, I did an analysis of all qualified relievers since 2002 and separated the 30 highest and 30 lowest groundball rates (Johnson was 24th).

	GB%	K%	BB%	BABIP	LOB%	Fbv	Fb%	HR%	HR Per Contact	WAR/60 IP	SD/MD
League	44.1	19.5	9.5	0.292	73.3	91.5	62.8	2.4	3.5	0.3	1.7
GB-Heavy	60.5	16.9	9.1	0.294	73.0	90.3	72.7	1.7	2.3	0.4	1.7
GB-Light	31.2	24.4	8.7	0.264	77.0	91.2	64.7	3.0	4.6	0.6	2.3

So not a whole lot of good things to say about the groundball heavy group. Jonny Venters was the only member of the group with a strikeout rate above 20%. They limit home runs pretty well, which is to be expected with so few fly balls. However, many of those groundballs are going for hits, while fly balls that aren’t leaving the yard are twice as likely to be outs.  That 30 point difference in BABIP is pretty huge, and that’s over a sample of more than 20,000 balls in play for each group. Overall, the decrease in home runs isn’t worth the extra hits and walks. With guys like Kenley Jansen and Rafael Soriano, it’s not surprising that the fly ball group features a much better ratio of shutdowns to meltdowns. For the most part, the groundball group is filled with situational guys that have bounced around with sporadic success. While relievers of all types tend to be unreliable, groundball and contact types are subject to the additional randomness of batted ball variance.

Seasons with inflated BABIP and ERA should be an expected consequence for a contact pitcher like Johnson. Of course, it would have been very difficult for the Orioles to demote Johnson to a lower-leverage bullpen role after the success he had in 2012. However, all signs indicate that Johnson is an average bullpen arm whose performance last season far outweighed his ability. He is better suited for the role he played in 2010 as a mid-leverage arm who was not limited to one inning. The Orioles should look for a strikeout arm for high-leverage situations. While Buck Showalter has consistently defended Johnson, not too many managers will bring back a closer after a season leading the league in blown saves.