Joba has been bad, but why are his peripherals so good?

Joba Chamberlain seems to be on his last legs, if not as a Yankee, then perhaps with the big league club for a while. For the first time since he made his debut in 2007, the team appears willing to deal the young pitcher. The ever-loyal Joe Girardi has even turned against him. He announced before and after Monday’s game that Chamberlain has lost the 8th inning job, his second demotion this season.

For the foreseeable future, the once off-limits Joba is now just a reserve arm in the bullpen, to be used when needed, ideally when the game is not on the line. This stunning fall from grace warrants a closer look to understand why Chamberlain is struggling, and what, if anything, can be done about it.

Unfortunately, Joba’s numbers don’t paint a clean picture. His ERA is 5.95, which is atrocious, but his FIP and xFIP are 3.01 and 3.41, respectively, both solid figures. What’s more is that he has improved his strikeout, walk and home run rates from 2009. In ’09 he struck out 7.61 batters, walked 4.35 and gave up 1.20 homers per 9 innings. This season all three figures have improved, standing at 9.99 strikeouts, 3.61 free passes and 0.64 homers per 9 innings. Any pitcher who improves his ability to strike batters out while walking fewer of them and giving up fewer home runs should be having an improved season, except for Chamberlain, whose 2009 ERA was 4.75, a full 1.20 runs better than this year.

The story of Chamberlain’s woes lies right where ERA and FIP diverge, in balls hit in play. Here a clearer picture of what’s going on begins to emerge. This season Joba is getting hammered, but the ball is staying in the park. His BABIP against has exploded from .320 in 2009 to .399 this year. Opposing players are hitting like Robinson Cano in April against Joba if they put the ball in play, which pretty much means if he doesn’t strike them out, because he has allowed only 3 home runs all season. Taking all at-bats into consideration, opposing players are hitting .299 against Joba.

Chamberlain may be getting unlucky (really, really unlucky). The discrepancy between his ERA and his peripherals is odd for any pitcher, but it is particularly odd for him. He’s pitched well in-line with his peripherals every other season he’s been a Yankee. If this is all due simply to luck, to the fact that balls hit in play against Joba are finding the gaps and not fielders’ gloves, then he is suffering from an especially long string of bad luck. The concept of the small sample size is abused and misunderstood in baseball’s statistical circles. Chamberlain’s 42.1 innings pitched on the season are, by definition, a large sample of innings because the statistical rule of thumb is to cap a small sample at 30 observations, something we had from Joba a while ago. Chamberlain may still be getting unlucky, but the data now points to a trend that defies the conventional tools even informed baseball fans use for analysis.

Joba struggled last season as well, but his problem was clear in 2009. Although his fastball is famous for its speed, it has never been a strong pitch. When Joba was at his best in 2007, according to Fangraphs his fastball was worth 1.51 runs for every 100 pitches versus 5.09 runs for his slider, his out pitch. In 2009, his fastball slowed to 92.5 mph on average, and was straight as a string, to quote Michael Kay. As a result it was actually a liability, worth -1.21 runs per 100 thrown in 2009, versus 1.30 runs for his slider. Chamberlain uses his fastball to set up his slider, but with the diminished velocity it was getting hit more.

Continuing the trend established previously, Chamberlain has actually improved on the effectiveness of his two main pitches this year, but to ill effect. His fastball velocity is up to 94.4 mph on average. Although it is still a liability, it is worth -0.75 runs per 100 thrown this year, a considerable improvement from last year. His slider is better as well, being worth 1.40 runs per 100 thrown this season (I am using values per 100 thrown to adjust for differences in innings pitched). 92.6% of the pitches Joba throws are fastballs and sliders, so he is essentially a two-pitch pitcher, and has been for his entire short career.

Although these numbers are frustrating, they help explain why the Yankees may have continued to give the ball to Joba for so long. If you remove his inflated ERA, he’s improved every aspect of his game this season. Unfortunately the results have been the exact opposite of what the team would expect.

One set of data does stand out that may help to explain Joba’s struggles. Opposing batters are swinging at pitches he throws outside of the strike zone 28.3% of the time, the highest rate since 2007, and they are swinging at pitches he throws in the zone 62.4% of the time, the highest rate in his career. This translates to an overall swing rate of 44.8%, which would be the highest of Joba’s career by far, if not for his 2007 campaign, when he got batters to chase pitches out of the zone 35.1% of the time, no-doubt fooling them with that nasty slider, and contributing to an overall swing rate of 47.2%.

The batters aren’t chasing. They’re hitting. Batters are making contact on 86.6% of the pitches in the zone they swing at, which is the 2nd-highest rate of Joba’s career, and they are making contact with 60.6% of the pitches outside of the zone that they swing at, far and away the highest clip of Joba’s career (his previous high was last season’s 55.9% rate).

I don’t believe that Joba is getting unlucky. It may be a baseball platitude to write this, but I watch the games. The hits he gives up may not be home runs, but they are laser beams the fielders can’t get to. From looking at this data my conclusion is that the league may have caught on to Joba’s diminished slider and straight fastball (Kyle Farnsworth had a plus-velocity fastball that wasn’t very effective also due to its straightness) and as result batters are swinging more. Joba is getting more swinging strikes this year, but that is only because last year players weren’t chasing outside the zone. This year they are, but they are often making contact, perhaps just enough to foul off his slider, if not crushing it. This may cause Joba, who remains a strikeout pitcher, to move away from his slider, his best pitch, and use his fastball or curveball to get the out. Of course, the problem is that both Joba’s fastball and curveball have been minus pitches.

Yankeeist still believes that Joba is too talented not to straighten himself out, but he clearly has some work to do to regain the trust of both the team and the fanbase.

6 thoughts on “Joba has been bad, but why are his peripherals so good?

  1. Hi Mike,Despite his current struggles, I feel the Yankees would need to get something pretty special back to trade Joba – and his value is pretty low right now, so i think any trade talk is all rumor. I feel the same way – every time he goes out there, I think of kyle farnsworth's tenure with the yankees. so, I agree, Joba isn't unlucky, he's just getting hammered.Does your analysis here shake your confidence in FIP at all?Fastball velocity and straightness: When everyone says 'fastball' I assume they mean a 4 seam fastball, which is traditionally a straight fastball and displays the pitcher's top velocity. since Joba started for most of the year in 2009, i was wondering if he was throwing the 2 seam fastball, which has that away-from-left-handed-batters movement (from right handed pitchers) with a slight dip in velocity – was he throwing a 2 seamer last year?when i was evaluating him, my approach was far simpler… i just added his hits (51) to his walks (17) and compared them to IP (42.1) and said to no one in particular, "wow, 68 is more than 42. no wonder he stinks." how do you feel about this method of evaluation? antiquated? ~jamie

  2. Trade talk may be rumors, but compared to previous years when Joba was strictly off-limits, this is sea change. I would trade Joba for Joakim Soria in a heartbeat, and I believe Brian Cashman would as well. I've never liked FIP as a statistic. I feel we put too much stock in it in the baseball stats community. Andy Pettitte and Mariano Rivera are two great pitchers who trust their defenses but whose FIPs suffer as a result. I like FIP and xFIP only early in the season to understand if a pitcher's ERA is justified. Once the season is a month or two old I've yet to see stats that I trust more than ERA and WHIP. My understanding is that Joba only throws a 4 seamer. Your approach is WHIP, which is a great stat for evaluating a pitcher's performance, as far as I'm concerned.

  3. Hey Jamie,I've seen a number of people call into question the reliability of FIP in light of Joba's recent struggles, but I have a feeling those doubting FIP aren't fully understanding the statistic.FIP is a measure of everything that a pitcher is specifically responsible for independent of his defense, i.e. strikeouts, walks, home runs and hit by pitches.As Mike pointed out, Joba's strikeout rate is up and walk rate is down, and he's giving up homers at half the rate he was last year. Hence the stellar FIP.But this is why, especially with pitching metrics, it's dangerous to rely on only one stat. While I like FIP a lot, it obviously doesn't tell the entire story.When batters are putting the ball in play (which has zero effect on FIP), Joba's getting crushed to the tune of a .399 BABIP against. That's actually the second-worst mark among all AL relievers, which shouldn't surprise anyone. The AL reliever with the worst BABIP against? Bobby Jenks. Rounding out the bottom 5 are Hideki Okajima, our own David Robertson and Mark Hendrickson. However, Jenks has still been the 11th-most valuable pitcher in the AL in terms of WAR. And believe it or not, Joba is the 16th.Even stranger, despite being the 16th-most valuable reliever in the AL, Joba also has the worst strand rate in the AL, at 61.1%, which shouldn't surprise anyone who's seen him pitch this season.What does all this mean? Well for one, Joba Chamberlain has been quite the enigma in 2010.I suppose the good news is, while Joba has obviously not gotten the job done in many respects this season, he's actually pitched fairly well within certain advanced metrics, and he really has no choice but to get better. Building his confidence back up with lower leverage innings, which it sounds like the Yankees are going to be doing, will probably be the best tack right now.We also have to remember that he's only 24 years old, and for more on that, make sure you check out Ben's excellent piece on Joba from earlier today.

  4. ha, that's what WHIP is? I'm a jerk. I'm not sure I understand what the value represents… Soria's WHIP is 1.200… i see he's had 48 base runners in 40 innings, which seems like a lot, but with 47 Ks, I can see how that can be effective, especially if the hits are scorchers like the gappers Joba gives up for extra bases…does Soria have the potential to start? i still think Joba can be a great starter – visions of last July still dance in my head…~jamie

  5. To Larry's point, Joba's season really is bizarre. If we ignore ERA, FIP and xFIP – that is to say any metric that attempts to measure how many runs he surrenders – he has improved in every category, including pitch effectiveness, except hits. I think the advanced metrics inflate Joba's value because he seems to produce either a strikeout, or a double. My hunch is that a certain amount of the value in WAR, for example, from a strikeout is fixed to represent, on average, what that kind of out is worth, which is typicaly more for a reliever than, say, a ground out, which has a lower value (I'm assuming). My guess is that these metrics assume that all pitchers will get a certain number of strike outs, ground outs, etc, and each represents a certain amount of value about the pitcher moving forward, strike out pitchers having the most value (this last part I'm certain of, as it pertains to advanced stats). Joba, however, isn't getting the other outs enough, so the value of his strikeouts is inflated. That's my hunch, anyway. WHIP is a great stat. It measures how many baserunners a pitcher allows, on average. Anything around 1.2 or lower is great (CC, Soria, Verlander). Anything under 1 is ridiculous (Mo). WHIP tells us a lot of things. It tells us, in general, how good a pitcher is because it measures all baserunners allowed except for HBP. It can tell us if a low ERA is luck (a lot of stats do this) because, say, a reliever has a great ERA, but a high WHIP, indicating that he will be clobbered soon, with runners on base. It can also tell us how efficient a starter is. A.J., for example, has a good career ERA but a high WHIP, and his innings pitched suffer as a result.

  6. hey guys,i'm an advanced stat idiot – i'm trying to improve, though. i tend to rely on raw data…when I like at WHIP and see 1.2 or less – i get that its good, but i dont know exactly what it means. when i do some quick addition and compare that to innings, i feel like i know exactly what to expectalways enjoy chattin with you guys!~jamie