A line that is commonly repeated this time of year is 'spring training stats don't matter'. You'll hear it from all corners of the internet and it's mostly true, but in the same way that Westley was mostly dead, meaning it's also slightly untrue. The fact is, right or wrong, jobs are won and lost on how players produce during spring training games. Case in point, Luis Mendoza.
If you'll recall Mendoza is competing for a rotation spot because Dayton Moore contracted Humberitis, a rare affliction that makes a general manager paranoid about losing a mediocre player because another mediocre player performed well for three months for another team (look it up!). Mendoza has pitched 6 scoreless innings this spring and now appears to be the favorite to be starter number five. In his last start he pitched 4 impressive innings, though it should be noted that the lineup the Rockies tossed at him wasn't going to strike fear in the hearts of too many pitchers (for example, career .186/.225/.289/39 OPS+ hitter Brandon Wood batted cleanup).
Danny Duffy was dominant in his first start but I think he's facing an uphill challenge now and may have to be damn near perfect the remainder of the spring. And even then it may not be enough, not as long as Mendoza is tossing up zeros. Sure you can argue that it's much too early for this sort of speculation but I'd counter with Yost and Moore are seeing exactly what they wanted, and what they expected. I mean, if they don't accept the general sabermetric view on Betancourt then you can pretty much assume they don't on Mendoza either. And of course I'm talking about the view that over the course of a full season his strikeout and walk rates would be a hindrance to sustained success. Recent history suggests this is, in fact, true.
If I'm right and he does make the rotation then I think we could see a repeat of what Sean O'Sullivan did last year. Like Mendoza, O'Sullivan is trying to survive with a low strikeout rate (4.2 SO/9). Last season he had a 2.16 ERA after his first 4 starts but his other numbers were telling a story that would produce an unhappy ending. In 25 innings he had 13 strikeouts and 13 walks and held hitters to an unsustainable .227 BAbip, meaning he was more lucky than good, and at some point the wheels would fall off. And boy, did they ever. His next 5 starts produced a 11.10 ERA and he was compassionately removed from the rotation making just one more start all year (a token September appearance in which he gave up 6 runs in 5 innings).
O'Sullivan provides a decent example on the difference between AAA and major league hitters. For Omaha last year he had 6.6 SO/9 and 1.9 BB/9 rates, numbers he came nowhere close to replicating for Kansas City (2.9 SO/9, 4.0 BB/9). So what Mendoza accomplished in AAA last year doesn't really matter because to succeed at the big league level he has to miss bats, and that my friends, is something he is unable to do.