When Rich Harden returns to the rotation on May 11th, for something between one pitch and the rest of the season, he will likely bump either Greg Smith or Dana Eveland from the rotation - and it is conceivable that at some point this season one of them could be marked for a stint in AAA.
But while it was Eveland, not Smith, who made the team out of spring training, and while overall Eveland has pitched pretty well, there are a number of reasons why I favor leaving Smith in the rotation and bumping Eveland into the "DiNardo role" (formerly known as the "Saarloos role" - which is long reliever, 6th starter, next guy sent to AAA, guy waiting for Harden to break so he can rejoin the rotation).
Going "inside the numbers," Smith has an edge that is actually substantial, in that Smith has allowed as many as 3 ER in only one of his six starts - and that was when he gave up a 3-run bomb in the first inning only to blank the Royals in the 2nd-5th innings - while Eveland has given up 3 ER or more in three of his seven starts. Another important difference: Smith has gone 6 innings or more in five of his six starts, while Eveland has failed to do so in four of his six starts - the last four prior to last night.
Going "outside the numbers," the differences between Smith and Eveland appear to be that Smith is already a "pitcher" - meaning he has control, command, poise, maturity, and a clear plan he can execute - while Eveland is still very much a "thrower" - meaning he has good stuff but has less of an idea of what he wants to do or how to do it efficiently or consistently.
As a result, Smith really has nothing to "work on" in the bullpen or minors, while Eveland might benefit from the extra bullpen work a long reliever can get, or from a stint in AAA. If one of them had to be optioned to AAA at some point, I would actually favor sending Eveland down even though it would use up an option year (#2), which it wouldn't with Smith (he uses up option #1 this year no matter what). Why? Because the team will still have an option on Eveland in 2009, and frankly if by 2010 he is scuffling enough that the A's want to send him down to the minors, that means he will have fallen far enough down on the A's "depth chart," probably behind Gio Gonzalez, Trevor Cahill, James Simmons, and Brett Anderson, and he will just be "the guy from the Haren trade who didn't pan out." By 2010, he needs to stick or move on.
Finally, while a sample of 6-7 starts each is small, the sample is larger than 6-7 starts. In April and May, Smith has pitched exactly as he did in the Cactus League and in his one start for Sacramento: throwing strikes, keeping hitters off balance, getting outs. In April and May, Eveland has also pitched exactly as he did in the Cactus League: Showing a good arm, putting a lot of guys on base, throwing a lot of pitches, getting out of more jams than not, and being inconsistent but promising.
Conclusion: Don't let the radar gun fool you. Eveland still has a lot to work on, while Smith just needs a lot of work - as in, every 5th day in the major leagues.