30 May 2013

O's Bullpen Is Not Overachieving Anymore

For a reminder of why baseball analysts often note that relievers are fickle, look no further than the difference between the 2012 and 2013 Orioles bullpens. Consisting of several of the same contributors, the 2013 bullpen has not been nearly as good as last season's surprisingly dominant group.

Sure, it's the end of May and the bullpen could turn things around the next few months, but it's going to be nearly impossible for them to approach last season's success. Take a look at the two seasons:


The biggest differences? Walks and home runs. In 2012, the O's had a HR/FB rate of 9.6 percent; in 2013, that rate has jumped to 12.1 percent. This season's combined BABIP (.280) is exactly the same as last season, but the bullpen is stranding about 4 percent fewer runners and inducing about 3 percent fewer ground balls. The slight bump in strikeouts is nice, but the increase in free passes and big flies adds up.

Last season, the team's top five relievers -- Darren O'Day, Jim Johnson, Troy Patton, Luis Ayala, and Pedro Strop -- outpitched their peripheral stats. O'Day, for example, was outstanding, but he had a high strand rate (85.1%). So did Patton (84.6%). Strop walked lots of batters but barely allowed any home runs (0.27 HR/9) and also had a high strand rate (83.2%).

But the outpitching peripherals thing, as you'd expect, isn't happening as much this year. O'Day is at it again (1.88 ERA/3.73 FIP/4.19 xFIP) with an even higher strand rate so far. Hunter (1.86 ERA/4.65 FIP/4.22 xFIP) is right there as well (100% strand rate). But (remember, small sample size alert) Johnson, Strop, and Patton have not pitched as well as last season and are the three biggest reasons for the bullpen's struggles. All three have walked more batters and have increased strand rates, and Johnson and Patton have a higher BABIP than last season. Strop's BABIP is only .216, but he has a ridiculous HR/FB rate of 27.3%. So that low BABIP doesn't mean as much when opposing batters are launching home run after home run.

Let's explore each of the three a little more below.

Jim Johnson:


Johnson isn't as good as he was last season, but he's not as bad as he's been this year, either. The jump in home runs allowed is a huge reason why he's been so bad, and if that number was around 9 or 10 percent, things wouldn't look as awful. Per Pitch F/X data, his four-seam (94.4 to 93.2) and two-seam fastball velocity (94.2 to 93.8) is down, but that hasn't affected opposing batters' plate discipline numbers all that much. They're actually swinging at the same number of pitches outside the zone (28%) and are even making less contact overall (83.3% in 2012; 81.9% in 2013). The biggest change appears to be that hitters are swinging at more pitches in the zone (from 55.7% to 61.5%) and are hitting Johnson's offerings harder (from a 16.4 line drive percentage to 20.5%). Perhaps he needs to work on changing speeds more often, or maybe mix his pitches a little better if his two-seamer isn't working as well.

Troy Patton:


Patton has been a lot worse across the board. Like Johnson, he's not as good as his 2012 numbers, but he's not this bad. At the very least, Patton has been able to get lefties out throughout his career (.272 wOBA against), but this year they have a combined .419 wOBA when facing him. That's not likely to continue, unless Patton keeps walking a bunch of batters like he's doing now. Patton's never been a hard thrower, but per Pitch F/X, his fastball velocity is down about 1 mph, from about 90 to 89. But unlike Johnson, opposing batters haven't been chasing Patton's pitches out of the zone as often. Last year, opposing hitters swung at 34.3% of Patton's pitches outside the zone. This season? 25.8%. They're also swinging less overall (50.3% to 43.3%) but are making much more contact (79.4% to 83.1%).

Pedro Strop:


Like most, I'm still not sure what to make of Strop. His arsenal of pitches is downright nasty, but he just can't seem to control them. His walk rate has jumped in each of his seasons in Baltimore -- from 2.19 to 5.02 to 7.13 -- and hitters are just laying off his pitches (32.1 O-Swing% in 2012 to 21.2% in 2013) because there's a good chance he's going to walk them before he can throw three strikes. Like Johnson and Patton, his fastball velocity is also down (96.4 to 95.3 (four-seam) and 96.9 to 95.6 (two-seam). If he can't throw strikes and get ahead of hitters, he may never get a real chance to approach that 2012 (or 2011) season again.


Triple R said...

1. O'day has been outperforming his perifs for his entire career (his career ERA is 119 points lower than his career xFIP) and I see no reason why that should stop now.

2. Johnson has been lucky for not one, but two years now, which may suggest a permanent trend; also, don't forget he had a very rough stretch last year after the All-Star break but bounced back to his normal lucky self after that.

3. Is TJ McFarland and his 8.39 K/9 for real? His K/9 was only 4.82 at AAA past year, and with Weaveresque velocity, that seems like the level he should e at in the bigs.

4. With Johnson and Strop, the perifs aren't all that much different from last year, but Patton is truly mindboggling. How does a guy who entered the year with a very good career K/BB of 3.64 suddenly start walking more batters than he strikes out? Just from watching him, his accuracy is horrible-Wieters will give him a target, and he'll miss by a foot or more. What gives?

5. (most importantly) Most relief pitchers will outperform their perifs over their career. Mo Rivera has a career ERA 79 points lower than his career xFIP. Trevor Hofffman has a career ERA 91 points lower than his career xFIP. Both of them have career BABIP's in the low .260's. Face facts-whether it's because of the low workload or just plain luck, relievers WILL outperform their perifs.

Matt Kremnitzer said...

You make some good points, and thanks for reading.

Yeah, looking at relievers is a lot different than starters. I'm not sure bringing up two of the greatest relievers, in Rivera and Hoffman, is really saying that much, though. None of the O's relievers are close to that level.

There might be too small of a sample to say much definitively. Anyway, I have more confidence in Johnson bouncing back than Strop or Patton. I just think he's the better pitcher.

As for McFarland, who knows? I guess we just need to see him pitch more.

Triple R said...

Another question (which I forgot to aks earlier) in regards to Brian Matusz: He'd striking out a lot more batters and walking a lot less than when he was a starter, but the major difference seems to be the ground balls--his GB% is a robust 53.3%, compared to 40.6% last year and 36.2% for his career. What has he changed to keep the ball down so much more, and is this common for a starter--turned--reliever?

Also, where does Chris Davis buy his roids? GNC doesn't seem to have 'em.