- FanGraphs Baseball - http://www.fangraphs.com/blogs -

Trying to Understand Bronson Arroyo

Bronson_Arroyo_2011For five years now, Bronson Arroyo has been better than his peripherals. Since 2009, only three pitchers have a bigger gap between their fielding independent numbers and their ERA, and those three didn’t come close to pitching as many innings. It’s tempting to say the free agent 36-year-old has figured something out… but what has he figured out, exactly? How has he become more than the sum of his parts? It has to be more than a whimsical leg kick.

Let’s use some basic peripherals to find comparable pitchers. His fastball struggles to break 90 mph, he doesn’t strike many out, and he doesn’t have great worm-burning stuff — but the control has been elite. Here are a few other pitchers that fit that sort of mold.

Name K% BB% GB% vFA (pfx) IP ERA FIP HR/9 BABIP LOB%
Bronson Arroyo 14.1% 5.5% 42.6% 87.6 1039 4.05 4.73 1.42 0.267 76.0%
Josh Tomlin 13.6% 4.5% 37.1% 88.8 329.2 4.72 4.54 1.37 0.273 68.6%
Jason Vargas 15.3% 6.7% 38.1% 87.4 837 4.10 4.33 1.14 0.280 73.3%
Dallas Braden 14.5% 6.2% 38.9% 87.6 347.1 3.63 3.77 0.73 0.283 71.5%
Bruce Chen 15.8% 7.2% 32.5% 87.6 613 4.42 4.60 1.28 0.284 72.8%
Kyle Lohse 15.0% 5.5% 41.5% 89.9 806.2 3.80 3.94 0.96 0.284 71.7%
Jamie Moyer 13.2% 5.8% 42.1% 80.5 308.2 5.22 5.27 1.66 0.284 69.1%
Randy Wolf 15.9% 7.7% 39.6% 88.7 791.2 4.07 4.46 1.13 0.285 74.5%
Jeff Karstens 13.8% 5.2% 42.8% 89.0 411.2 4.15 4.33 1.18 0.288 72.1%
Eric Stults 14.8% 6.2% 39.6% 87.6 345.1 3.81 3.73 0.73 0.291 71.0%
Freddy Garcia 15.2% 6.7% 39.8% 87.4 515.1 4.54 4.58 1.31 0.292 72.6%
Rodrigo Lopez 13.1% 6.4% 38.2% 88.1 315.1 4.74 5.17 1.63 0.297 70.2%
Kevin Millwood 15.3% 7.7% 41.3% 89.9 604.2 4.30 4.54 1.16 0.297 72.6%
Livan Hernandez 12.8% 7.2% 40.7% 84.7 570.2 4.48 4.11 0.80 0.307 69.1%
Average 14.5% 6.3% 39.6% 87.5 559 4.29 4.44 1.18 0.287 71.8%

There, even among his peers in terms of walks, strikeouts, velocity and grounders, Arroyo stands out. He has the biggest gap between his ERA and his FIP, based perhaps on the largest strand rate of the group, and the lowest batting average on balls in play.

Maybe a look at Arroyo’s history would give us a clue. He used to strike guys out and show batting averages on balls in play above .300 even. Has he altered his pitching mix since 2009? Thanks to BrooksBaseball, we can check:

Pitch Type Before 09 After 09
4 Seamer 29% 17%
Sinker 17% 28%
Changeup 16% 20%
Curve 36% 32%
Cutter 2% 3%

It looks like Arroyo has ditched the four-seamer some, for more sinkers and changeups. That can be the way of older pitchers — fastball percentage and age are negatively correlated, albeit weakly (.102 r^2) — and at first blush, it could be the full reason for the lower BABIPs. Changeups do have lower BABIPS than other pitches (since 2009):

Pitch Type BABIP
FT 0.314
SI 0.310
FA 0.307
FF 0.304
FC 0.296
CU 0.295
SL 0.294
FS 0.292
KC 0.289
CH 0.287
KN 0.284

Unfortunately, you’ll notice sinkers aren’t known for their batting average on balls in play, so any added sinker usage probably offsets his added changeups. And though his ground-ball rate has been better since the change, it hardly seems enough to explain this phenomenon. He went from allowing .95 ground balls per fly ball to upping that to 1.1. And the extra ground balls haven’t helped him suppress home runs — his homer rate has only gotten worse as he’s aged.

He’s been with one team, and they’ve had similar personnel around the infield? There could be something to this. The Reds have allowed the lowest batting average on balls in play in the National League since 2009. Brandon Phillips has good glove, the team has usually focused on defense at shortstop, and among Joey Votto‘s many strengths is being above-average in the field. And that batting average on balls in play has probably helped Arroyo strand a few extra runners over the course of the last five years.

Bronson Arroyo has separated himself from pitchers with similar skill sets by stranding more runners and allowing fewer hits on balls in play. He’s altered his pitching mix some, and that might be a small part of the picture. More importantly, perhaps, he’s been with the same team for five years. If he doesn’t return to Cincinnati, his new team will be betting that his stuff will play up in front of a new infield, and that the magic continues.

Thanks to wiki commons user Bubba Fan for the header image and Jack Moore for running the correlation between fastball percentage and velocity.