Of course, he allowed three runs over five innings yet still lowered his ERA to 11.79, so "best-ever" for Shields is a one-size-fits-most tag.
As you might expect from the line -- 5 IP, 5 H, 3 R, 3 ER, 4 BB, 3 K, 0 HR -- it doesn’t answer a whole lot of questions. He addressed one problem in that he threw first-pitch strikes to 15 of 23 hitters. He worked ahead in the count far more often, setting up his changeup and curve more than he had in previous starts. Alex Avila said Shields corrected himself quicker, too:
"He was playing more in the strike zone early on in the at-bat than in previous starts," catcher Alex Avila said. "Throughout the innings that he pitched, there were times where he kind of got out of himself a little bit and rushed a little bit, but he was able to make the adjustment much quicker than he did in his previous outings."
That describes the fluctuating amount of command Shields showed throughout his start. But as I said in the recap, he still looked Mat Latos-ish to me, and velocity was a part of it. He didn’t hold it that well:
Match the audible grunts (I think Fenway Park’s acoustics amplify them) against the diminishing radar-gun readings, and it still felt like a labored start for a sizable portion of the afternoon. Perhaps that isn't the aftertaste had he not walked the two batters he faced in the sixth with zero quality strikes thrown.
That said, after three starts that had few historical parallels for all the wrong reasons, he had to start somewhere, and this counts, especially against one of baseball’s most dangerous offenses. If he can’t sustain it against Minnesota next week, the alarms remain within arm’s reach.
After watching Tim Anderson strike out twice in as many bases-loaded plate appearances against Craig Kimbrel, I wondered if the problem with Anderson batting leadoff isn’t so much his aversion to walks, but his lack of exposure to MLB high-leverage relief pitchers.
It’s suboptimal that Anderson is 60 plate appearances without a walk, and hasn’t seen a three-ball count in 25 of his last 26 PAs, but everybody knew that coming in.
Thinking about the damage Anderson had done, most of the examples I could generate occurred in the first half of the game. His splits on Baseball-Reference.com jibe with that notion, with his inning, starter/reliever and power/finesse splits all favoring early-game success.
In particular, Anderson is 4-for-18 with nine strikeouts against relief pitchers, and 1-for-9 with six strikeouts against power pitchers. The rest of the league doesn’t hit Kimbrel or his ilk well, so I’m not surprised to see Anderson look overmatched in his first looks against this grade of reliever. However, a lot of these power-pitcher matchups contribute to the extra plate appearances that top-of-the-order hitters accumulate over the course of the season, and the Sox are seeing a lot of them lately due to the amount of close games.
Even knowing the splits and Thursday’s results, it’s not entirely instructive and prescriptive. For one, we're dividing an already-small sample, and all it takes is one or two good games for Anderson to flip this on its head.
Usual caveat aside, putting Anderson at the bottom of the order wouldn’t spare him from tough situations, as ninth-hitter J.B. Shuck preceded Anderson and made weak contact both times. Batting order doesn’t dictate late-inning clutch situations. If it did, perhaps White Sox fans wouldn’t have hated Tyler Flowers so much.
We haven’t seen enough to make sweeping statements, but we’ve seen enough to be on the lookout for more similar situations to see how Anderson adjusts. As long as the Sox keep playing nail-biters, Anderson should get more chances against more of these arms. This increases the incline of the learning curve a little, and the Sox are counting on his reputation as a quick study.