Kevin Towers, Zach Duke, and Pitch Data

If you paid attention to the Pirates last year, you probably know lefty Zach Duke had a pretty rough go on the mound. He posted a career-worst 5.72 ERA, and his 4.95 FIP tied for another career-worst. The Pirates traded him for Cesar Valdez, who himself is considered to be a rather fringy arm, in the offseason.

With his new organization, Duke is expected to win a starting rotation job. Diamondbacks GM Kevin Towers recently spoke to MLB.com about his reasons for acquiring the 27-year-old:

"“Sometimes a change of scenery is good,” Towers said. “It kind of re-energizes you and gives you a chance to clean the slate and start anew. I like left-handed pitching, I like athletes. His stuff really didn’t deteriorate over the last couple years. He had a down year last year, but if he had a good year we wouldn’t have had an opportunity to get him.”"

A lot of that makes sense.Yes, a change of scenery can help. Athletic lefties are good to have, and Towers is right that Duke would’ve been tougher to get if he were coming off a good season.

But there’s one problem.

Zach Duke‘s stuff hasn’t deteriorated over the last couple of years? Obviously, the idea of a pitcher’s “stuff” is somewhat subjective, but let’s take a look at the most objective data we have on his pitches, the Fangraphs pitch type and pitch value charts:

It’s difficult to see anything that didn’t deteriorate in 2010 compared to 2008-09. You could make an argument for the curve, since slower curves aren’t necessarily worse, and his 2010 curve was more effective than it was in 2007-08 (but not 2009).

Beyond that, though? Let’s list the deterioration, just from 2009 to 2010.

1.) Duke lost 1.4 mph on his fastball.
2.) Duke lost 2.5 mph on his slider.
3.) Duke only lost 0.4 mph on his changeup, therefore losing a full 1 mph of separation between it and the fastball.
4.) Duke’s slider was easily the least effective it ever was.
5.) Duke’s changeup was easily the least effective it ever was.
6.) Duke’s fastball had its second-worst season, and worst since 2007 (You could also say the slider was worse in 2007 than 2010, but he almost never threw it in 2007).

I found all that data in about thirty seconds. It’s not exactly difficult to come to the “he’s lost some stuff” conclusion from it. But there’s Kevin Towers, an MLB GM, who disagrees.

This isn’t to say Towers is necessarily wrong. After all, Duke is just 27, and he certainly could recover 1-2 mph if he got on a throwing program that worked, a la Barry Zito last year. And if Towers knows about his velocity and effectiveness losses, and thinks Duke can get back to 2009 form anyway, I’ve got no problem with that, even if I don’t necessarily agree.

But I wonder if he even knows. Kevin Towers is an insider, and insiders should certainly have more sources than I do. The mistake I feel some of these executives can make is relying solely on inside sources. For example, let’s say Arizona scouted Duke, and their scout reported that he threw in the upper-80’s, with an 80-ish slider, slow curve, and low-80’s change. That’s all completely correct, but just from reading that, one wouldn’t get the impression that Duke’s stuff has changed.

But in dealing with MLB pitchers, there’s obviously a whole host of freely available data that everyone should make sure to consult. Eyes can be deceiving, so we should always check them up against something more objective. It’s easy to gloss over a 1.5 mph difference from one year to the next when watching a pitcher, because if a pitcher threw 88 and now throws 86.5, there are still going to be lots of 86s, 87s, and 88s on the scoreboard. The radar guns, though, are going to pick up the difference over the course of the season.

From what little one can discern from Towers’ comments, he may well be unaware of what the objective data says about a decently important acquisition, which is pretty shocking to me. Perhaps he knows something I don’t, or perhaps he was simply using overly politically correct rhetoric to back up his move–after all, it’d be kind of weird to say “Yeah, Duke’s just a junkballer now”–but it just seems very odd for an MLB GM to make a statement that seems to fly in the face of very concrete data. It’s not like Dayton Moore arguing that UZR is wrong and that Yuniesky Betancourt is a great defender. We’re talking about average velocity and pretty clear-cut pitch effectiveness stuff here–nobody’s going to say, “No, Duke actually throws 90-92; the data’s wrong.” It’s just baffling that such an important executive could be uninformed about something so basic.