I began with David Aardsma for alphabetical reasons. Shawn Kelley followed for the same reasons. Jumping to Brandon League might make some think that Mark Lowe is up next. Those people have decent pattern recognition skills. He's not though. You should have realized that Mark Lowe didn't appear enough to warrant mentioning and got traded so screw him. Anyways, Brandon League.
The knee jerk answer if you have been following this blog or even FanGraphs for awhile is to blame the pitch selection. As I first mentioned both here and FanGraphs (and was subsequently repeated numerous times by many other people), League's splitter was the most best pitch in baseball last year as defined by the percentage of pitches resulting in a swing and miss. It was a pitch he started throwing only last season and we had high hopes that he would continue with that step forward and all those strikeouts coupled with some regression from his BABIP (.322), strand rate (just 68%) and HR/FB (15%) would reveal Brandon League as the next breakout reliever.
It hasn't worked out that way even though those who follow only ERA will think he's done fine this season. Truth is, he has been decent. We just hoped for more than decent and frankly, League is capable of delivering on those hopes.
The knee jerk answer above isn't the full answer but it isn't incorrect either. In 2009, League threw one splitter pitch for every 2.3 fastballs. That rate has dropped in 2010 to one splitter for each 5.5 fastballs. League is featuring his fastball more often and it is hurting his ability to get strikeouts as he is right back to the level he was accustomed to in Toronto pre-2009. He hasn't lost the ability to get those back though. Looking at the results for each pitch offers confirmation of this:
2009 Fastballs - 5% whiff rate, 66% strike rate, 51% ground ball rate
2010 Fastballs - 7% whiff rate, 65% strike rate, 59% ground ball rate
2009 Spliiter - 35% whiff rate, 59% strike rate, 61% ground ball rate^
2010 Splitter - 29% whiff rate, 47% strike rate, 40% ground ball rate^
^stupidly small samples
League is throwing fewer splitters and hitters are swinging at them less often, down four points from 2009. That has suppressed League's swinging strike rate and consequently eliminated the extra strikeouts he used to record.
It is not all bad news though. Increased usage of the fastball has predictably resulted in fewer walks and his net walk rate is down two points and even his grounders have increased. Neither have moved enough in the right direction to offset the loss of strikeouts however.
Returning to the old pitch mix would offer an almost immediate salve to League's strikeout woes. Under the assumption that his effectiveness with each pitch would remain the same as they have been in 2010, raising his splitter throw rate to 2009 levels would bump his swinging strike rate up to 13.5, nearly a match for 2009's 14.0%. Those extra missed bats would inevitably lead to more punch outs (and a few more walks).
The other part of the answer to what has gone wrong with Brandon League has to do with the expected regressions mentioned in the second paragraph. The BABIP has regressed as expected. The strand rate has just gotten worse, but since League has lost a small dinghy's worth of strikeouts he no longer is expected to post an above average strand rate. His current rate (67%) is still a bit too low, but that is not the chief culprit. That would be the home runs.
Let me get this out of the way again. I have looked at it a couple times with different methods and data and I, and others, have yet to find any solid and compelling evidence that ground ball pitchers suffers from higher HR/FB rates. So it is with confidence that I say that League's 18% HR/FB rate is too high. It's way too high and getting that down to a more normal range would go quite a ways to propelling League into that 'good' territory of reliever.
In conclusion, throw more splitter Brandon. You can control that part. Don't sweat the home runs though, those should take care of themselves.