For the first month of the season, the Red Sox didn’t just have a bad bullpen, they had one of the worst bullpens in recent memory. Manager Bobby Valentine‘s relief corps pitched to a 6.10 ERA with a 5.13 FIP in April, both easily the worst marks in the game. Boston’s pen has gone on to post a 2.31 ERA with a 2.97 FIP since the calendar flipped to May, bringing their overall season performance down to a much more respectable 3.55 ERA and 3.75 FIP.
A turn around like that can be attributed to many things, first and foremost just simple regression to the mean. A .337 BABIP and 16.4% HR/FB ratio certainly aren’t performances you’d expect an entire pitching staff sustain over a full season. It’s possible, just unlikely. Secondly, the Red Sox did what you would expect them to do and made some personnel changes. Mark Melancon (49.50 ERA/37.55 FIP (!!!)) was shipped to the minors while Justin Thomas (7.71/3.27) and Michael Bowden (3.00/6.38) were cut loose. Changing the names is the easiest way to change performance.
Among the replacements was former Miguel Cabrera trade bait Andrew Miller, who is finally starting to pitch like the sixth overall pick in the country is expected to pitch. Granted, he’s doing it out of the bullpen rather than out of the rotation, but it’s better than what he had been doing. Anyway, Miller has pitched to a 2.13 ERA (2.42 FIP) in his limited action (just 12.2 IP) and has been death on lefties, holding them to a .193 wOBA with 32.1 K% and 3.6 BB%. One reason for his success is the return of his fastball velocity, which now regularly bumps the mid-90s…
Velocity is one thing, but Miller has also dropped his arm angle ever so slightly. The plots below are Miller’s 2011 release points on the left and his 2012 release points on the right. Click the images for a larger view — or play with the graphs yourself — and you’ll see that he’s releasing the ball anywhere from 3-6 inches lower — and further away from the center of the plate horizontally — than he has in the past.
The lower arm slot has helped Miller raise the whiffs-per-swing rate on his breaking ball to north of 40% after sitting mostly in the low-to-mid-30% range in the past. He isn’t finding the strike zone that much more than he had coming into the season — 52.8 Zone% in 2012 vs. 51.0% from ’07-’11 — but part of the lure of power guys like Miller is that they don’t have to be in the zone to be effective. Just being near the zone is enough to generate bad swings and weak contact, at least in theory. In a past, he wasn’t even close to the zone.
Miller’s track record as an effective pitcher is about a dozen innings long, so it’s tough to give him the benefit of the doubt and think that it has finally clicked. For a while last year it looked like Dontrelle Willis was back to being effective and that was just a fluke. Ian Snell is another enigma we can point to. The good news is that we have some tangible evidence for Miller’s improvement, specifically the velocity spike and the lower arm slot. This is also his first year as a full-time reliever. The effective version of Andrew Miller is one of baseball’s many white whales, and whether the success is sustainable or not, he has been a big reason why the Red Sox have been able to right their bullpen ship over the last six weeks or so.