What's wrong with Phil Coke? In a nutshell: He isn't missing bats. That might be a symptom, or it might be a cause. That part is a bit harder to say. All we can say for certain is that his swing, contact and whiff rates from 2011 are consistently out of whack with those from the rest of his major league career.
Initially, I thought of Phil Coke as not one, but four pitchers, so I could compare how he may have approached the game differently. For instance, he pitched in relief all but the final game of 2010. During that season, batters swung and missed about 12% of the time (per Baseball-Reference). He also had relatively even splits from right-handed and left-handed batters. As you might expect, the left-handed Coke did better against lefties. But he didn't do bad against either. The Tigers' decision to give him a chance in the rotation wasn't a bad one. He looked like he could handle a back-of-the-rotation assignment.
That leads me to look at the next three "versions" of Coke. The first one was a starting pitcher for most of the early season until May 23. This one had an ERA in the mid-3s and was a pretty effective pitcher. He had a quality start in four of his first eight games, falling an out short of a fifth quality start. How unhappy can you be about a 50% (nearly 62.5%) quality start percentage from your fifth starter? The ninth start of the season was a problem. Playing in wet conditions, Coke slipped when going to cover first base. He hurt his ankle. That game he left in the fourth inning with what he felt like was a minor injury. The next day he hit the disabled list. His ERA as a starter at that point was 3.62. His swinging strike rate was about 8%.
Coke returned to the rotation almost exactly 15 days after hitting the disabled list. Did he return from his injury too soon? Did the layoff affect his mechanics? Did something mental happen? I don't know the answer. (I'll look closer later in the story, though.) All I can tell you is that he wasn't the same pitcher. In his next five starts, he gave up four or more runs in four of them. His ERA was 7.20 during that span. His swinging strike rate was down to 6%. After he gave up 14 runs across two starts, the Tigers could not justify keeping him in the rotation any longer.
Which leads to the final version of Coke: back in the bullpen. Initially this looked like it would turn out well for the Tigers. He struck out 3 batters in an inning on the mound on July 4, though he did allow a run. Since then, the story has not been as happy. Coke has pitched in 11 games, and opponents have an OPS of .939 off him. They're hitting line drives in 20% of their appearances against him, and swinging for strikes at just 5% of his pitches. In fact, in the 11 innings since July 4 he only has 3 strikeouts. Certainly his ERA has improved -- to the mid-4s during that span -- but things don't seem to be going all that well on the whole.
The more I thought of it, the less I liked the "four Cokes" theory. There seemed to be two things going on. The first is that he certainly changed his tactics when he became a starting pitcher. He took a little bit of velocity off the ball, as he was expected to stick around for six or seven innings, not one. He also decided to pitch to contact.
"You don't need to worry about striking everybody if you're pitching to contact. And I haven't thrown many pitches, man. I think I'll be just fine [in September]."
Here's a chart that summarizes how batters have reacted to Coke's pitching from 2009 through today, validating that.
|Year||Zone%||Out of zone
|Out of zone
A few things jump out to me. The first is that Phil Coke is putting the ball in the strike zone nearly half the time, a jump from 42.4% last year. But looking at the chart 2010 may have been the year that was out of line.
Second, as I said, batters are making contact a lot more often. The whiff rate was double digits in 2009 and 2010 but has fallen to 6.5% this year. The cause for that appears to be the success batters are finding on balls outside the zone. They are making contact nearly 80% of the time on the pitches they swing at out of the zone. By the way, they're fishing less frequently.
Change in pitching strategy and possibly some fluky results aside, the bigger issue to me is that the date of the injury seems to be an incredible dividing line in results.
His OPS against in 2010 was .669. Before he was injured in 2011 was .661. After the injury, as I mentioned, it was in the 900s. Breaking it down further, his OBP against in 2010 was 350. It was .306 before he was injured this year. It has been .416 against since the injury. Part of that is walking batters. He issued 16 walks in 217 batters faced before the injury. He has issued 18 more in just 180 batters since. Slugging, too, has seen a rise. It was .348 against in 2010, .355 before the injury. It has been .416 since. Basically, it's been a nightmare for Coke the past two years. If he puts it outside the strike zone, bad things happen. If he puts it inside the strike zone, worse things happen.
The final stats worth noting indicate Coke's fastball has not been as effective this year. Using weighting to value pitches, we learn that Coke's fastball has been at least one run per 100 pitches better than the average pitcher from his first appearances in 2008 through the end of 2010. However, in 2011 his fastball is valued at -1.40 per 100 pitches. That's a swing of 2.50 from last year. Looking at PitchF/X results helps explain it.
So far this season he has thrown his four-seamer 59% of the time, 64% for strikes. Batters swing at it 59% of the time and whiff 3%. They put it in play 26% of the time per Texas Leaguers. Compare that to 2010. He threw it about 58.5% of the time, 69% for strikes. Batters swung 52.7% of the time and whiffed 7%. They put it in play a bit more than 22% of the time. (TL) Taking our glance at the fastball a step further, batters have whiffed on it just 2.5% of the time since he returned from the injury while putting it in play almost 30%.
Putting it all together, I can't help but come back to the ankle injury. On May 23 he slipped and reportedly turned his ankle while fielding a bunt. He went to the ground in pain and left the game at that point. Later, on crutches after an MRI, he told reporters it was a bone bruise caused when two bones in the ankle separated and came back together. That doesn't sound like the type of injury that should be lingering of course.
But why might an ankle injury be a big deal? It was to Coke's "plant foot." Every time he throws a pitch, he lands on that leg with a rather large amount of force as his arm continues its motion. Pitching is such a highly technical action, especially at the highest levels, that it's easy to understand how a little pain in a key place could mess up the entire production. In the MLB, little flaws turn into big problems. Now it should be noted that Coke's velocity hasn't suffered. But the ball hasn't been going where he wants it to go either. So it seems possible his mechanics are off a little bit after the injury. That could be due to some pain or weakness in the ankle. Maybe it's some flaw that arose during the layoff. No matter the reason, it seems like the most likely scenario.
As for a projection for the rest of the season, I feel like things have been going a bit better for him lately. He has cut his walk rate back to normal levels since returning to the bullpen. He did have a run of five scoreless games in July. Some of the results he's been seeing seem fluky. I still feel like there's some pitching strategy and mechanical issues that have to be addressed, but believe worries about Coke need to be kept at reasonable levels. It wouldn't surprise me if he turned a corner at some point by the end of the year. I just can't say with any certainty when that point might be.