Ever since Danny Duffy's wondrous 2 hit performance against the Orioles nearly a month and a half ago, many RR contributors have been sitting on pins and needles awaiting his disastrous crash back to earth. However, while his reply to the almost-perfect consisted of two disasters in a row against the Angels and Astros, Duffy has, in terms of actual results, continued to be very, very good. Many still conjecture that his continued success will only lead to an even harder fall, but what if the "new and improved Duffy" cited by Goldberg and FSKC/KC Star homers is not actually a mirage?
Let's look at the numbers. Ever since entering the rotation on May 3, Duffy has sported a 2.66 ERA, 6.16 K/9, and 3.64 BB/9, combining for a 1.69 K/BB. He's shown a mild decrease in walks, but the increase in control has also lead to a decrease in strikeouts - in the end, causing a worrisome 4.00 FIP. These last few stats certainly are cause for concern. Some have cited a decrease in velocity as the cause for the increased control and decreased K%, but his average fastball during those starts has been 93.4, correlating closely with his career velocity of 93.7. (Note that all of these career statistics include Duffy's 2014, so the differences are, in reality, even more pronounced than I am reporting here).
Let's dig even deeper. Duffy's SwStr% has plummeted from a career mark of 7.8% to 5.9%, and this has correlated to an increase in Contact% - from 82.5% to 87.1%. None of that looks particularly rosy.
However, pure contact rates are not very telling, especially when it would seem prudent for Duffy to make use of one of the best defenses in the league. So, an important question to ask is, What type of contact are hitters making? Encouragingly, Duffy is allowing hitters to square him up less often, as his LD% has gone down a full 3%, and both FB% and IFFB% have increased by multiple percentage points. Many have said Duffy is enjoying a joy ride upon the BABIP Fairy, and that probably explains at least some of Duffy's success this year, but a decrease in LD% also explains part of the lower BABIP. Nevertheless, I do expect more balls to start bouncing in the grass as the season continues, but maybe not as much as some people foresee.
An increase in fly-balls can be good, but it often accompanies a corresponding increase in home runs. However, Duffy's HR/FB has always been better than league average, which probably has some to do with his home park and some to do with the way he pitches. Additionally, while the Royals' defense as a whole is one of the best in the league, the outfield defense is absolutely premier, so increasing the amount fly balls is actually probably a good strategy, especially if he can continue to induce the type that don't go over fence.
Last step - pitch type. Duffy has always been a fastball, change-up, curve pitcher. The fastball has the best stuff, but the curve can be devastating, especially when he's throwing it for strikes. One big change this year is that Duffy finally realized that his change up isn't that good and often got battered when he left it up in the zone, so he's now throwing it less often, using it more for show than for results. As those over at KofK have noted, it will be important for Duffy to eventually develop a 3rd pitch, as getting through the lineup three times is much more likely if he can keep hitters guessing. It is possible though, as Rex and Ryan like to point out, that Duffy has replaced the change up by changing fastball speeds more often. I don't have the expertise to corroborate this assertion with data, but it does seem as if it may be true. Unfortunately, average pitch velocities are not a good way to analyze this possible change in approach.
The projections don't like Duffy for the rest of 2014 and are unconvinced that he'll continue to prevent runs as well as he has in the first half. While I agree that he will likely see an increase in his ERA over July and August, I think the data explained above may give us reason to be encouraged and the increase won't be as drastic as ZiPS and Steamer predict.