Write the sentence, "The Rangers won the AL pennant because of a strong bullpen," on an index card, mail it back to July 31st, and it wouldn't be much of a spoiler. The Rangers actually making it through two playoff rounds, somehow beating the juggernaut Red Sox? Slight spoiler. But everything the Rangers did at the trade deadline had to do with building a strong bullpen for the playoffs.
Exactly how it happened, though, is a little surprising. The Rangers didn't root through the bargain bin -- they paid full retail for a couple of highly respected late-inning guys. Mike Adams cost the Rangers two pretty good prospects (Joseph Weiland and Robert Erlin), but he pitched as well as expected.
Pretty clear that Koji Uehara could be bumped from roster if Mark Lowe is healthy.
A couple months ago, Uehara was on the short list of most reliable setup men in baseball. About 75 days later, he's in danger of being left off a World Series roster entirely.
If Jon Daniels' goal was to build an unhittable bullpen so the Rangers could get away with their starting rotation averaging five innings per start, he's the smartest man alive. It's worked so far. But Uehara has been a dirigible accident in the playoffs. His 13.50 ERA in the ALCS is actually an improvement over his ∞ ERA from the ALDS. He's allowed three home runs and five earned runs in just 1⅓ postseason innings. He's looked so bad, it's hard to remember him being any good in the first place.
But he was good. He was extraordinary in Baltimore, but since arriving in Texas, he's allowed eight homers in 18⅓ innings. The Rangers have been putting him in lower and lower leverage situations, and that strategy has reached its ultimate conclusion: Ain't a lower leverage situation than not being on the roster.
Uehara is the ultimate test in one's belief in sample size. Everything he'd done as a major leaguer up until two months ago indicated he was more than a quality reliever -- he was one of the best. But after 18⅓ bad innings, the Rangers aren't sure if he's one of the seven best relievers in their organization at the moment.
It's not a velocity thing. His velocity has actually been going up since he came into the league.
It's probably not a control thing. He only allowed a single walk in the regular season, though he's walked three in the postseason so far -- just two fewer than he walked in all of 2010.
If the Rangers do leave Uehara off the roster, it would be to put Mark Lowe on the roster. Lowe is coming off a hamstring injury, and he's struggled mightily with control throughout his career. He's also allowed more home runs in 45 innings with with Texas this year (6) than Uehara did in his 44 innings with Baltimore (5).
This isn't to suggest that the Rangers should put Uehara on the roster. Maybe there are mechanical flaws they're picking up on, or something that suggests that his decline isn't just a fluke. They'd probably know better, and that organization is run well enough to get the benefit of the doubt.
But the situation highlights just how hard it is to truly evaluate a reliever. Eighteen innings is nothing. In Cliff Lee's last 18 playoff innings, he's allowed 26 hits and 15 runs. That means absolutely nothing when predicting his future worth. Uehara's 18 innings all happened within a two-month stretch, so they might have predictive value. Or they might be as uninformative as Lee's last 18 innings in the postseason. I have no idea.
The Rangers think they do. Or, rather, they don't want to risk the chance that Uehara really is broken. It's a sound strategy, but it could be overly cautious. Texas could be leaving one of the better relievers in the game off the roster.
Of course, he'd still be around the fourth-best reliever on the Rangers even if he were at full capacity. Must be nice.