Was Derek Lowe’s Implosion Predictable?

The Derek Lowe experiment is over. Speculation was rampant that Lowe was about to lose his rotation spot after he got shelled for seven runs in 2.1 innings Tuesday night and Josh Tomlin was moved to the bullpen, but now it’s official: Lowe has been designated for assignment, which in all likelihood means he will never pitch in an Indians uniform again.

We’ve seen a large number of “I told you so”‘s around the internet over the last couple days as Tribe fans who were skeptical of the team’s deal for Lowe last fall expressed their vindication. But such sentiment is misguided. There’s no denying that the 39-year-old veteran has been absolutely abysmal on the mound—he’s 2-8 with an 8.80 ERA, a 2.12 WHIP, and a .913 OPS against in his last 12 starts—but there’s no way anyone could have known that things would get this bad.


It’s important to first distinguish between two easily conflated but very different predictions one might have made about Lowe: That he would not keep up his torrid early-season pace, and that he would cease to be a serviceable MLB starter. I’m not here to discuss the former. No reasonable analyst would think that a pitcher in his late 30’s could maintain the Cy Young-like 2.15 ERA he held at the end of May, especially since he was barely striking out a batter every four innings. Even if you don’t buy into the sabermetric defensive-neutral ERA estimators that suggested Lowe was getting lucky, the concept of regression to the mean—i.e., that a performance like Lowe’s was unsustainable—is not (or at least, shouldn’t be) controversial.

Expecting Lowe to come back down to earth was only rational, and I know many Indians fans and writers weren’t very high on him before the season. What I’m referring to is the idea that this complete and utter collapse was predictable. As a member of the Cleveland media tweeted Wednesday morning:

Who could have possibly seen Derek Lowe’s implosion coming? I mean, aside from everyone who saw his stats last season.

There’s no getting around the fact that Lowe’s 2011 numbers were ugly. He lost an MLB-high 17 games for the Braves last year while posting a 5.05 ERA (his 77 ERA+ was his worst since 1997), including a 6.20 ERA and an .850 OPS against after the All-Star Break. There’s a reason the Braves were willing to eat $10 million of his 2012 salary to get him off their hands.

But, as we noted at the time of the trade, Lowe’s struggles were largely the result of bad luck. The DIPS estimators FIP, xFIP, and SIERA all put his true-talent ERA between 3.65 and 3.75. The problem was his .327 batting average on balls in play, which tied the highest he’s been stuck with in a decade despite the game’s recent trend towards pitching, and his uncharacteristically low 66-percent strand rate. Many attributed these symptoms to Lowe’s alleged tendency to give up hard contact, but while Lowe has some history of underperforming his peripherals there is nothing on his lengthy résumé to justify that assertion.

But forget sabermetric theory for a moment. Ignore the fact that Lowe’s peripheral numbers remained virtually unchanged during the stretch run last year, and that he has a long track record of pitching far more effectively. Even if we assume that the 14 starts he made in the second half of 2011 comprise a big enough sample size to assess his true talent level, there’s a huge difference between an ERA around 6.00 and an ERA approaching 9.00, as Lowe’s has in his last 12 outings. Lowe’s slump makes his numbers last year look like Walter Johnson‘s.


And even if it had been clear that Lowe still had farther to fall, his problems now are completely unrelated to those he faced last year. Lowe’s kryptonite last year was the number of batted balls that fell for hits. Call it luck, call it an inability to induce weak contact—it was probably a little of both. But whatever the cause, his .366 second-half BABIP was to blame for his late-season collapse.

Lowe’s given up an unusually high number of batted-ball hits this year too (.333 BABIP), and anyone who’s seen him pitch knows that the issue isn’t just luck. But there’s a bigger problem with how he’s pitched this year: He’s not striking anyone out. His 2012 K/9 rate sits at a meager 3.1, and he’s inducing swinging strikes at a rate of less than 1 every 18 pitches. It’s virtually impossible for a pitcher with numbers like that to enjoy any sort of sustained success—how are you going to get batters out if they know you’ll give them something to hit?

This weakness came completely out of nowhere. Lowe lost any pretense of strikeout chops when he moved the rotation full-time in 2002, but before this year his K/9 rate had never slipped below 5.0. In both 2010 and 2011, his strikeout rate was more than double what it is in 2012. So even if you had made the irrational pre-season prediction that Lowe would lose his roster spot out of pure ineptitude, you would have been basing it on a problem that didn’t exist yet.

It’s no surprise that Lowe has regressed after his phenomenal April and May, and many Tribe fans had been anticipating disappointment from him for about nine months now. But it’s misleading to point to Lowe’s slump in 2011 and declare that his severe decline was inevitable, and I don’t think anyone honestly predicted that Lowe would suffer such a drastic downfall.