John Sickels warns to watch out for small sample sizes before getting too excited about Zach Duke. Specifically, Sickels points to the fact that Duke's strikeout rate is higher in the majors than it was earlier this year at Indianapolis.
As I began writing this post, I was going to suggest that Duke actually improved as the minor league season went on. Now, I'm not sure that's true. It is true that his ERA improved. At the end of April, Duke's ERA at Indy was 4.15, and his strikeout rates were unimpressive. He struck out eleven Toledo Mud Hens May 3, then got hit hard in his next start against Columbus to raise his ERA to 4.17. But he didn't have a bad start after that, and nine starts later he had lowered his ERA to 2.92.
After May 8, though, Duke pitched 67 innings and struck out 40 batters while walking 15. That's a fine ratio and it will keep runs off the board if, like Duke, you don't allow home runs. But it comes out to less than five and a half strikeouts per nine innings, which is indeed well below Duke's strikeout rate in the majors so far.
Then again, we all know to beware of small sample sizes. We also know that Duke isn't likely to keep his ERA below 2; no one thinks that Duke is superhuman. Duke could lose a strikeout every nine innings and still be well above average for a big league starter. His performance spike since arriving in the majors probably isn't real. (I remember someone once joking of Tike Redman's remarkable half-season with the Pirates in 2003 that 'If he learned something, he seems to have learned it on the flight from Nashville.') But if Duke regresses the rest of the year by posting an ERA in the twos or low threes, I think I can be happy with that.