With Eduardo Rodriguez here to stay, who leaves Boston's rotation?

"if he goes to Pawtucket, it's on his way to Baltimore." John Farrell left no room for interpretation following Eduardo Rodriguez' second major league start: with two great performances in two tries, the young lefty is in Boston's rotation to stay.

The question now is how the Red Sox make room for him. Yes, they've already technically done so. Rodriguez has been on the roster for a week now, and they could keep right on throwing out a six-man rotation so long as the starters keep giving them as many innings as they have of late. But Farrell himself has said he doesn't like six-man rotations, with this exception largely in place to handle a three week stretch without a scheduled day off.

This week's postponed game and the subsequent doubleheader don't throw a wrench in the plans, either. The streak will end on Monday, which is exactly when the rotation will come back around to the portion disrupted by the rain. On Tuesday, they'll be able to roll straight through on regular rest. But given Farrell's preferences, they will very likely do so with one fewer starter.

Clay Buchholz obviously has his spot locked in. He's been excellent for the Red Sox of late, and has managed to bring his full season numbers into more than respectable territory even after that one disaster against New York.

Rick Porcello is sporting a 5.01 ERA after a couple rough outings, and still has the home run specter floating over him. But he turned in eight strong innings last night, and still seems to just be dealing with the one problem that is as likely to sort itself out as not, with his peripherals remaining fairly strong.

A little bit ago, Wade Miley might have been a prime candidate. But even after a mediocre performance to end the month (with the defense partly to blame), he produced a 3.49 ERA in May, and really seems to be finding his rhythm in the post-Nieves era. It's possible that a real disaster could have the Sox looking to put him on the disabled list to figure some things out, but barring that, his spot also seems safe.

That leaves two options: Steven Wright and Joe Kelly.

Wright, to be fair, has done nothing to earn consideration. In three starts for the Red Sox, he's allowed just seven runs in 17 innings of work. He's often come out of the gates looking shaky, or destined for disaster, but the knuckleball hasn't really let him down yet. He's gone out there and essentially removed the pitching from the equation, leaving it up to the offense--a good night will earn them a win, a bad night the loss. Granted, right now that means a loss more often than not, but you can't blame Wright for that.

Kelly is pretty much the exact opposite. He's allowed two earned runs or less in three of his last five starts, and six or more in the other two. In fact, Joe Kelly has yet to allow three or four earned runs in any of his 10 starts this season. It's always either a gem or a disaster. There is no in-between. Unfortunately, taken as a whole, it also produces a 5.83 ERA, which is just not really acceptable in this day and age. Or even at the height of the steroid era, to be honest.

Amusingly enough, the way this Red Sox team is hitting, they probably actually stand a better chance to win on any given night with Joe Kelly on the mound than Steven Wright. This team doesn't seem to win anytime it allows four runs, which would make six innings and three runs from Wright and three innings with one run from the bullpen a loss. Those rare Joe Kelly gems are more likely to produce a win than Wright's consistently solid outings just because of how consistently cold the lineup is.

But that's not a reason to keep Joe Kelly around over Steven Wright. After all, the Red Sox can't be basing their decision on how to win the most games with their lineup producing at dead-ball era levels, because so long as it is, the team isn't going anywhere. There's no point playing to maximize your wins in a situation where you can't possibly make October. For Farrell and Cherington and everyone else involved, they have to make their decision based on the assumption that the lineup will get better since, if it doesn't, they may as well put Pablo Sandoval on the mound for all the difference it will make.

Instead, the reason to keep Kelly around over Wright is that sending down Steven Wright to Triple-A or moving him into a long relief role is no big deal. Doing the same to Kelly, however, is something of a statement. The acknowledgment that there's something seriously wrong with the idea of Joe Kelly, Starting Pitcher, and that something has to change, be it his role, or something drastic about the way he approaches that role.

Right now it feels like all of Boston is just waiting for that one catalyst. That one final straw to push the 5.83 ERA up over 6.00 and force the Red Sox to pull the trigger. Kelly thwarted expectations by surviving his last start. He's got one more left before the Sox are likely to make a choice, as does Steven Wright. Chances are that, by inertia alone, it's Joe Kelly's spot to lose.

But if you've been paying attention, you know that there's pretty decent odds that he does just that and loses it.