The Year James Shields Was Different

Three winters ago, we got into a lot of arguments about James Shields. He was at the center of a very polarizing trade and people took sides. You remember it, so I won’t rehash things other than to remark on how funny it is that the James Shields-Wil Myers blockbuster has actually become the Wade Davis trade. Wade Davis! The guy who gave up 5.92 runs per nine in the season following the deal.

Life’s little insanities aside, Shields was very good for the Royals during his two seasons in Kansas City. He was worth 4.0 and 3.3 WAR, respectively, and helped push them over the hump and back into relevance. Would they have gotten there without him? It’s entirely possible, but he was a key player on the team during their renaissance and deserves some recognition for it. You will note, however, that Shields signed elsewhere after the Royals lost the 2014 World Series and then the team won the 2015 title without him.

One of Shields’ hallmarks, and one of the main reasons the Royals acquired him, was his consistency. You were pretty much assured more than 200 innings of good, non-elite run prevention and above-average fielding independent numbers. Shields was as predictable as a person could be in baseball. Then he signed with the Padres.

The basics are easy enough to review. In Shields’ nine major league seasons entering 2015, he had a park-adjusted ERA worse than league average just twice, in 2006 (108) and 2010 (131). He had a park-adjusted FIP worse than league average once, in 2010 (104). His only season of fewer than 200 innings was his first season, in which he made 21 starts and tossed 124.2 innings. The 33-year-old was sure to age, but there weren’t any warning signs that suggested his decline would be unusual.

Fast-forward one year and Shields is coming off the worst fielding independent season of his career. In 202.1 innings, Shields recorded a 116 FIP- to go along with his 105 ERA-, which was the third worst mark of his career by that measure. But that’s not super interesting on its own. You can file “33-year-old Pitcher Has Worst Season” in the cabinet marked “Not News.” The interesting part is that Shields’ fall didn’t manifest itself in the way you would expect.

If you hadn’t seen any Shields news all year and I told you that he had a bad season relative to his baseline, your number one inkling would be to consult his BABIP and the potentially disastrous Padres defense. That would explain the ERA jump given that he had just spent two seasons in front of the Royals glovemen, but it wouldn’t speak to the FIP spike. So your next thought would be HR/FB%, at which point you’d remember he moved to Petco and would assume that couldn’t be it. So you’d think he lost something on his pitches and the strikeout rate fell. That’s the rote analysis version of what leads to most single-season pitcher declines.

Maybe I’ve sold your imagination short, but if your assumptions were high BABIP, fewer strikeouts, and no homer effect, you’d be wrong on all three counts. James Shields had his worst season in 2015 and it came about in a very strange way.

Let’s get the easy one out of the way first. James Shields allowed a .299 BABIP in 2015. In 2014, it was .295. In 2013, it was .298. In 2012, it was .292. His career BABIP allowed is .298 and you have to go back to 2010-2011 to find his outlier seasons (one high and one low). So a BABIP-based explanation doesn’t hold water. It’s not as simple as saying that some balls just happened to fall in 2015 that normally wouldn’t.

If you’re used to the way FanGraphs lists stats, your brain probably has muscle memory when checking in on a player. Your eyes float to the right of the page to look at WAR. When you noticed that Shields’ was low in 2015, your eyes bounced over and saw his bad FIP and below average ERA. Then you found BABIP, which provided no answers, leading you to look across the board from strikeout rate to walk rate to home run rate. And that’s where it really gets confusing.

Shields had his worst season (maybe second worst season depending on your philosophical bent) while posting the highest strikeout rate of his career. And while facing the pitcher in the NL helps, even if you exclude all 62 of his plate appearances against the opposing pitcher, he still had the highest strikeout rate of his career at 23.8%. If you remove the other pitchers, the mark only narrowly beats his 2011 and 2012 strikeout rates, but it remains significantly about his 2013 and 2014 seasons any way you slice it.

And the strikeout rate was based on a lot more swings and misses. His Contact% dropped 6.9 percentage points from 2014 to 2015, one of the biggest drops among pitchers who threw 100 innings in each season. That contact rate change went along with far few pitches in the strikezone (a six-point drop according to PITCHf/x). He expanded the zone and got a lot more swinging strikes as a result.

But there was a cost to that, which was his increased walk rate. Despite walking just three of the 62 pitchers he faced in 2015, his overall walk rate was a career high 9.4%. His previous high-water mark was 7.2% and had only topped 6.0% on four occasions. If you compare his strikeout and walk rates from 2015 to his career norms, the 2015 move to higher strikeouts and higher walks costs him something like 0.05 worth of FIP. That’s basically nothing.

His home run rate, on the other hand, explains a whole lot more. Shields allowed 1.47 HR/9 in 2015 which is much higher than his career 1.11 HR/9 and 0.91 HR/9 from 2014. Home run rates can be finicky. They don’t always indicate anything other than what happened. Shields has allowed a career HR/FB% of 11.7%, and it was 8.6% and 9.7% during his two years in Kansas City. In 2015, at Petco Park, it was 17.6%!

If you trimmed that HR/FB% to 12%, which isn’t even as far as xFIP takes it, you can shave 0.70 runs or so off his 2015 FIP. In other words, James Shields had a bad season because he managed to give up a lot of home runs while calling Petco Park home.

In fact, he allowed home runs at Petco Park at a nearly historic level. Entering the season, Chris Young’s 2006 1.89 HR/9 at Petco was the most for a Padres pitcher who threw 50 or more innings in San Diego in a single season. Shields finished the year allowing 1.74 HR/9 at home in 2015. That would be more impressive if Ian Kennedy didn’t manage to allow 2.00 HR/9 at Petco this year as well.

In 2015, Shields pitched in a way that generated more strikeouts, more walks, and way more home runs than we’ve seen from him in the past. And while the strikeout and walk trade wouldn’t be a big deal on its own, pairing it with the home run spike led to disastrous results.

One thing that jumps out is that he lost more than a mile per hour off his fastball from 2012-2014 to 2015. Maybe that explains the home runs, as it could serve as a proxy for worse stuff overall, but it doesn’t really line up with the extra strikeouts.

Another thing that jumps out is the sliding release point, moving toward third base.

That movement to the third base side of the rubber also lines up with the approach he took to the strikezone. Compare 2013/2014 to 2015:

It’s hard to miss the degree to which Shields pitched away from the center of the zone in 2015 relative to his time in Kansas City. If you split it up by handedness, he typically worked away to lefites and down to righties, but the shift away from the zone is pronounced in both cases. This certainly doesn’t seem like an accident.

Expanding the zone in this manner lines up well with more strikeouts and more walks, but it doesn’t totally explain the home run issue. Losing a tick on the fastball might fit the narrative there, but if batters could crush his weakened arsenal, why would they swing through it so much?

We probably have to chalk at least some of the home run problem up to one of those strange baseball oddities. No one has a true talent HR/FB% near 18%. Let’s stipulate that the strikeout and walk changes were the result of an intentional shift in approach and the home runs were part fluke and part unintended consequence.

But that leaves us with one major question: why would the eternally consistent and successful Shields move to an easier league and perfectly friendly park and totally reinvent himself? I can think of four potential, non-mutually exclusive reasons.

First, perhaps the Padres, either through the front office, the field staff, or the catchers, convinced Shields that attacking the zone wasn’t the right move for him in 2015. It wouldn’t be shocking for a team to have a philosophy like that. We know some teams preach certain pitch types, while teams like the Pirates believe in pitching inside. Maybe Shields was going along with what his new club thought was best.

Second, it’s possible Shields knew he didn’t have the same velocity on his pitches and decided he couldn’t live in the middle of the zone without getting clobbered.

Third, it’s possible that Shields arrived in San Diego to find a defense that was substantially worse than the defense he had known in Kansas City. Not every Padre was a problem in the field, but the club was something like -30 runs on defense in 2015 compared to the otherworldly Royals. Maybe Shields was afraid to attack the zone because he wanted to cut down on balls in play. If that was the goal, he succeeded nicely.

Finally, Shields might be feeling the effects of fatigue and/or injury. He’s had a big workload in his career and hasn’t ever broken down. Maybe that’s catching up with him and the drifting to the third base side of the mound is really an indication that he doesn’t quite have the shoulder or elbow health needed to slide into his normal arm slot. If that’s the case, missing the zone down and to the arm side might simply be accidental.

All of the options make some sense. The idea that he might be afraid of his defense is the most interesting, but history tells us adjusting to lesser stuff or being hurt is probably more likely. Steamer projects he’ll return to form in 2016, but Steamer also can’t know if he’s injured.

The Padres signed Shields for less than Shields was seeking last winter, but $75 million requires something like 10 WAR if you want to get fair value for your investment. After a 1 WAR season in 2015, that looks like a tall order. Shields could bounce back, hit his Steamer projection, and everyone could live happily ever after, but it requires you to believe the home run problem won’t continue to some extent and the strikeout/walk change isn’t a harbinger of decline.

The Padres are in a tough spot because they could probably stand to take some time to rebuild. That’s a problem because Shields won’t fetch much in a trade until he shows 2015 wasn’t a sign of the end, meaning that he’s simply going to be a financial burden during a retooling period if they want to go that route.

But that might offer an opening for another team that’s on the 2016 playoff bubble. There’s a decent chance that Shields still is a 3 WAR pitcher. He’s at the nadir of his trade value but if the reason for the down year was a style choice rather than a sign of physical wear, he’s a good bounce back candidate if you put him in a new environment. Teams might not want Shields at 3/$65M, but I would gamble on him if the Padres were willing to eat $15 million for the chance to grab a useful prospect.

He has an opt out after 2016, but I’m not sure he would use it unless he has a really strong season. If that happens, you can enjoy the fruits of the performance and take a potential draft pick for your trouble (assuming the new CBA doesn’t eliminate that entirely). There’s risk, of course, but unless you see red flags in his medical reports, the odds that he’s a good pitcher again seem strong enough for it to make sense.

James Shields was traded three years ago because the Royals wanted a sure thing. He might prove to be an interesting target this winter for precisely the opposite reason. His uncertainly coming off his worst season will make him easier to acquire, and I think the team that takes that risk might find themselves with cost-effective mid-rotation starter.