16 August 2013

What's Wrong With Jim Johnson?

Photo: Keith Allison

The short answer?  Nothing.  Jim Johnson is who we thought he was (sorry for the link, I couldn’t resist).  He’s a groundball pitcher who pitches to contact and doesn’t strike out a lot of batters.  His tremendous success in 2012 has almost been a double-edged sword, as his club record 51 saves gave some people unrealistic expectations that he was an elite reliever; someone who is nearly automatic in save situations.  However, despite his success in 2012, Johnson should not be viewed in that light.  Typically, when you think of an elite reliever, you think of someone who can dominate hitters, and limit the number of batters who reach base (i.e., high strikeout rate/low walk rate).  Jim Johnson doesn’t really fit that bill. 

To illustrate this point, let’s take a look at the 2013 saves leader board.

When you think of the best relief pitchers in the league, many of the names above come to mind, and for the most part, they all share similar traits as high strikeout/low walk pitchers.  Only 6 of these top 15 pitchers have a strikeout rate less than 25%.  Of the ones who don’t, Mariano Rivera, Addison Reed, and Steve Cishek are fairly close.  Edward Mujica is just over 20%, but his BB% sits at an incredibly low 1%.  Rafael Soriano has the lowest strikeout rate on this list, however, a 17.7% strikeout rate is unusual for Soriano, who has a career rate of 25.4%.  Then there’s Johnson, who is having one of the best strikeout seasons of his career, with a rate nearly 3% higher than last year.  Still, among all relievers with at least 30 innings pitched, Johnson only ranks 136th in K% and 65th in BB%. To put it another way, if you could choose to have any one of these relievers, would Jim Johnson even be in the top 10, even though he leads the league in saves?  Probably not.

Don’t get me wrong, Johnson is a very good pitcher who induces a lot of ground balls (also a good trait to have in a reliever), but those ground balls can find holes and are often times subject to the whims of the Luck Dragon.  Other than the number of blown saves, Jim Johnson’s 2013 doesn’t look all that different from his 2012.  His strikeouts are up slightly, but so are his walks.  His LOB% is virtually the same, and while he’s getting less ground balls than he did from 2011-2012, his 55% GB% is only slightly under his career level of 57.3%, and still well above the league average of 44.1%.  

There are differences between this year and last year though, and unfortunately they are the result of the aforementioned dragon.  In 2012, Johnson sported a well below league average BABIP of .251, helping fuel his 2.49 ERA.  Additionally, the few times that he allowed fly balls, they only left the park 6.8% of the time.  So far in 2013, his luck has shifted, as both his BABIP and HR/FB% have jumped well above his career levels, to .321 and 10.3% respectively.  The fact that Johnson’s xFIP is only 0.23 runs higher than it was in 2012 helps confirm that his luck has changed considerably in comparison to last year.

Even if there appears to be sufficient evidence, it’s difficult to say that only a pitcher’s luck has changed and leave it at that.  If you can point to a more tangible reason why Johnson may be struggling, you can look at the movement and command of his sinker, his most effective pitch and the one he throws most often (48.8% in 2013).  According to Pitch F/X information from Brooks Baseball, Johnson’s sinker has had less horizontal movement this year than it did in 2012, which when combined with a slightly lower velocity, will likely help the pitch become much more hittable. 

Additionally, Johnson’s command of his sinker has not been executed as well this year either, as the two figures below will show.  As you can see, in 2012 Johnson peppered the outside edges of the strikezone, rarely leaving the pitch out over the plate.  Compare that to the current season, where he’s throwing the pitch over the middle of the zone with increased frequency.  

Jim Johnson is a good relief pitcher, and one that the Orioles should be happy to have on their roster.  But he’s not an elite relief pitcher, and just because he racked up 51 saves last year does not make him one.  His underlying numbers tell us that he benefitted from good luck in 2012, the way those same numbers tell us he is suffering from bad luck in 2013. Additionally, decreased effectiveness and command of his sinker may also be contributing to his struggles.  

Despite Johnson’s recent string of blown saves, the Orioles don’t really have another option who would present an obvious upgrade, though they may be able to find more success finishing games by shifting to a match-up based approach.  However, managers are not generally fans of using a “closer by committee”, so the Orioles will likely stick with Johnson for the moment and pray that the Luck Dragon has a change of heart, which isn’t all that bad of an idea.


Anonymous said...

Insanity is doing the same thing and expecting different results. While I understand that "closer by committee" is often considered a last resort, it's probably the least evil when there's no dominant reliever available. At the very least, you would get to spread out the risk by diversifying that way. A possibility is this:

* Use Hunter for 2-inning saves;
* For one-inning saves, use O'Day vs. R and Matusz vs. L.

Unknown said...

that would definitely be a possible solution, but what are the chances they even try it?

steve said...

I understand the caveat that pitch/fx classification type should be treated with caution, but JJ relies heavily on his 2 seam fastball. He threw it about 60% last year and 46% this year. His pitch/fx pitch value for this pitch has dropped from 6.8 to -3.9 this year. I'm assuming that's a severe drop in quality but I'm not sure how to interpret these numbers exactly so please correct me if I'm wrong Jon. The drop in quality in his bread and butter pitch most likely explain the greater % of HRs given up as well as his increase in BABIP this year (.321) in comparison to the past two years (about .260).

Jon Shepherd said...


Two things I would note: (1) pitch value is not a statement of the quality of the pitch, but a description of how effectively the pitch has been used which can be a quality issue or a use issue or a small sample size and (2) I am unsure how to really use pitch values effectively because I am not sure whether they ever really stabilize in their worth. For instance, a two seamer can be bad because other pitches are not working. It is a complicated issue, so I tend to avoid those numbers.

steve said...

Gotcha. Very informative. Thanks Jon!