"The Red Sox bullpen has turned it around, allowing just seven runs in its last 39 1/3 innings over 11 games. The relievers have a 2.80 ERA over the last 39 games, lowering their collective ERA from 5.42 to 3.86."
Predictably enough, the very day the Boston Globe prints the passage above, the pen blows a game. Hideki Okajima pitched a third of an inning, giving up two walks, two hits, and three runs. Delcarmen allowed one of Oki's runners to score (sac fly) and Hansen finished it off with a 3-run ninth.
Okajima picked up his 6th blown save (the BS statistic). A cursory glance at his stats gives us this:
These are pretty awesome numbers, especially considering that in these 28 innings he's only given up 2 home runs. Whence the 6 BS, then? Well, Okajima has allowed 11 of 14 inherited runners to score, a shocking increase from last year. In 2007, Oki only allowed 4 out of 28 inherited runners to score. While Okajima's peripheral stats (BB, HR, Ks) are still strong, he simply hasn't been as effective this year in keeping down opposing clubs.
Some people, mostly fans of other teams, claimed last year that Oki's success stemmed from his funky delivery, and that as opposing batters became familiar with it, they would clobber him. The evidence doesn't really bear this out, however.
Looking more closely at the circumstances of Oki's runners trouble, we see a bunch of high-leverage situations. He came in three times with the bases loaded (vs. LAA 4/24, DET 5/5, and BAL 5/14), and allowed 8 of those 9 runners to score. Those three games account for nearly all of Oki's inherited runners. Anecdotally, holding on inherited runners seems like an unstable area for many relievers. For example, Cleveland's Rafael Betancourt allowed only 3 of 33 to score last year; this season he's allowed 4 of 7 to score (and been MUCH worse than Oki, to boot).
So far in his 28 innings, Oki has only given up 9 runs, all earned. You'd expect more runs if the league were catching up to him, obviously, but you'd also expect them to be better distributed. Of those runs, 8 were scored by the Orioles.
We're not even halfway through the season, so it's possible that the rest of the league, or at least our division, will start hitting Okajima. However, I'm encouraged by his stats, particularly his high K-rate (8.04/9 innings), and believe he will be a top reliever for the rest of the year. Oki's 2008 might be worse than 2007, but not by much, if what he's done so far is any indication.
If the trials with inherited runners and/or Baltimore continue, then Francona will likely turn to other relievers for those situations. Lopez and Papelbon are both good at suppressing inherited runners. Moreover, on the off-chance that Oki's production falls off a cliff, the team can look for set-up men both internally and externally.