Has James Shields Simply Lost It?


In 2010, James Shields had a disaster of a season, going just 13-15 with a 5.18 ERA (although a 4.24 FIP). Then he broke out in 2012 with a 16-12 record, a 2.82 ERA and a 3.42 FIP. Now he has struggled to a 4.44 ERA and 3.92 FIP in 2012. This chart is pretty scary.

Not exactly a perfect ABBA, but the ERA’s certainly are, and on a whole 2012 has looked a whole lot more like 2010 for Shields as opposed to 2011. What is going on here?

Walks and home runs stem from similar problems: control, the ability to throw strikes, and command, the ability to place the ball in the area of the zone that you want it (e.g. high or low, outside or inside). Let’s look at some stats that can help us get a better idea of what Shields’ control and command have been like the past few years. (For simplicity’s sake and to get the biggest possible sample size, we’ll look at 2012 as a whole.)

Control isn’t simply about strikeouts and walks. For example, sometimes a strikeout can come only after missing wildly with several pitches out of the zone, burning a pitcher’s pitch count. So in addition to walk rate, a good stat to look at is the percentage of pitches thrown within the strike zone. The league average is around 49%. In both 2010 and 2011, Shields’ zone percentage was at 51.4%. This year, he’s down to 46.0%. That won’t do him any good. But throwing less pitches in the zone isn’t always a bad thing. Shields is forcing hitters to swing on his pitches out of the zone at a 34.2% clip, well above the 27.5% league average and his 31.8% and 32.6% rates the past two seasons. But throwing that many pitches out of the zone could mean something entirely different- that Shields has needed to throw more offspeed pitches out of the zone because his fastball has been ineffective. We’ll confirm that later when we look at the Pitch F/X data.

For command, we’ll look at a couple different things, groundball percentage and line drive percentage. If Shields has been able to keep the ball down, he should have a high groundball rate, and that is indeed the case as he has a 53.6% groundball rate, the highest of his career and well above the league average of 43.5%. He managed a 46.2% groundball rate in 2011 and 41.3% in 2010. That is a good sign, but it may go along with the higher percentage of breaking pitches that he has thrown, and then there’s the matter of the Rays’ defense, which has not exactly been spotless this season, getting Shields into even more trouble. Then there’s the matter of line drives. It has been proven (by this Baseball Prospectus article for example) that there isn’t a high correlation for line drive percentage from season to season. Anecdotally, one reason that could be the case is that line drives are most of the time caused by leaving pitches over the heart of the plate and a pitcher’s command will fluctuate from game to game and season to season. In any event, line drives are a decent measure of how much a pitcher has left the ball over the middle of the plate over the course of a season. In 2012, Shields has posted a 17% line drive rate compared to the 18% league average, and it’s the same mark that he posted in 2011. In 2010, he was at just 22%, and that explained a lot of his struggles. But wait a second- if Shields has left about the same portion of pithces over the heart of the plate in 2o12 as he did in 2011, why has he allowed so many more home runs? Line drives are different than flyballs, but the home run flyballs are the same type of thing because they’re hit about as hard. That’s where xFIP comes in, calculating FIP replacing home runs allowed with the expected number of home runs allowed given that one of every ten flyballs allowed should be a home run. By that measure, Shields has a 3.47 mark compared to his 3.92 FIP, which resembles his 3.42 xFIP from 2011 (he had a 3.25 xFIP that season). But the higher number may once again stem from the fact that he has thrown more offspeed pitches, and when you miss up with offspeed pitches, they’ll go a long way.

In summary based on the stats, Shields’ struggles have something to do with bad luck and poor defense, but a lot to do with a big dependence on offspeed pitches as well. To check out the validity of the second part of that statement, let’s look at Shields’ Pitch F/X data from the past three seasons, courtesy of Brooks Baseball, in my original displays.

We have seen that parts of Shields’ season have been similar to his terrible 2010, but other parts similar to his great 2011. Let’s put his Pitch F/X data from the three seasons side-by-side and see what we can figure out.

Look first at the key. We can see that Shields’ use of his four-seam fastball (in blue on the graphs) has decreased each season since 2010, from 35% to 30% all the way down to 19%. In 2010 and 2011, it was the pitch he used the most- in 2012 it’s in a virtual tie for third. Shields’ use of all his fastballs (four-seam, cutter, sinker) has gone down each season as well, 61% to 52% to 49%, not even the majority of his pitches. He’s used his changeup more each year until it became his most-used pitch this season. That has gotten Shields into major trouble, as we have talked about above. His hitters are seeing his offspeed pitches more, getting more comfortable with them and not getting fooled as often.

Now look at the lines of the three graphs. The thing that stands out the most is the difference between the blue line, his fastball and the green line, his changeup. From 2010 to 2012, the lines get more and more similar until in 2012, they’re virtual indistinguishable until the end of the blue line. Shields’ feel for the arm slot of his already superb changeup has just gotten better and better. But that hasn’t helped him so much- his ERA and FIP have gone back up in 2012. Why? Because he’s throwing his fastball less and less. It doesn’t mean a thing to sell your changeup as the four-seam fastball you don’t throw too much. Shields’ fastball has never been a great pitch for him, but it has come to a tipping point in 2012 as hitters are sitting on it and hitting a crazy amount of home runs off of it. It has gotten increasingly exploited the more the league has seen Shields. But now shift your focus to his changeup on all three graphs and in the keys. Shields is throwing his changeup 1 MPH harder than he has in the past, and the result is somewhat less movement on the graph. Shields needs all the movement he can get right now, and the 1 MPH isn’t enough of a difference to warrant that. And then look at his cutter, which he has thrown more and at much higher velocities in 2012, but with much worse movement than ever. He also throws it from a slightly different arm slot than his four-seamer and changeup, making his changeup a little less effective. And top it all off, Shields has lost his arm slot on his curveball a little bit, making it more of a groundball pitch than a swing-and-miss pitch. Shields’ repertoire is trending downward all around. Can he recover?

Shields getting back to anywhere near his 2011 levels will have to do with him reestablishing his fastball. He has to either get more movement on his four-seam or improve his cutter to the point that it can be a consistent plus pitch. This is something that Shields can do. He’s a veteran willing to make the necessary adjustments. He’s going to find some way to turn his 2012 season around. But don’t think that Shields is the true ace that he appeared to be in 2011. He has one devastating pitch, his changeup, and a good group of pitches after that, but nothing clearly plus. His control is still great and his command is as good as ever. That’s the difference between Shields being the pitcher he is and an enigmatic fourth or fifth starter. Shields has the ability to turn these bad trends around and be a good number two starter in the big leagues moving forward. If he can get the right mix of his pitches and tighten a few things up, he will still be a very good pitcher. When James Shields is on, he can carry this team. Shields is out of it right now. There’s no two ways about that. But if they have any hope of contending this season, the Rays need Shields to start looking like the pitcher we know he can be, and fast.