It has been quite the Major League debut for Astros starting pitcher Jarred Cosart. Through 8 starts, he has allowed just 9 earned runs over 51 innings for a minuscule 1.59 ERA. The 23-year-old rookie is no doubt making waves as one of the better young pitchers in the game. And while that may well be true, it certainly isn’t because of how he has performed thus far.
It is not often that you find a pitcher who has walked more batters than he has struck out. But, that’s exactly what Cosart has done, walking 26 batters, while whiffing just 25. In fact, since 2004, only 40 starting pitchers who have thrown at least 50 innings have actually walked more batters than they have struck out. Their weighted average ERA was an unsurprising 5.56 versus a 4.24 population (n = 1,768) average. So how on Earth has Cosart allowed so few runs?
Well to start, his BABIP sits at just .250. That’s rather surprising though because heading into last night’s game, he was an extreme ground ball pitcher, allowed a league average rate of line drives and induced a below average rate of pop-ups. His batted ball distribution would suggest an above league average BABIP if anything. He also pitches in front of the Astros defense who rank 28th in UZR/150. Furthermore, the team’s pitchers have allowed an aggregate BABIP of .305, which ranks as the third highest in baseball. So any way you look at it, the Astros defense has been poor this season, yet Cosart has managed to suppress hits when the ball has been put into play against him.
But the low BABIP isn’t even the half of it. His LOB% is a crazy 86.5%. So when Cosart is putting runners on, he’s stranding them at a higher rate than every other qualified pitcher in baseball. Oddly, his skills actually have been better with runners on than with the bases empty. His strikeout percentage has increased with runners on, and then again with runners in scoring position. Similarly, he has basically cut his walk rate in half with men on base. This is of course a rare skill. Pitchers normally pitch better with the bases empty, otherwise they would just pitch from the stretch all the time.
Since it doesn’t logically make sense that a pitcher would be able to sustain such type of bases empty/runners on performance splits, we have to assume it’s just a fluke. Of course, you’ll always hear the clichés about a pitcher “bearing down” with runners on, increasing their focus and blah blah blah. That’s just an example of developing a story to explain a statistical quirk, because people hate to use the term luck. That’s likely all it is though. If the pitcher had the ability to increase his focus with runners on, why not do it with no one on as well?
Last, Cosart has displayed an amazing ability to prevent the long ball. With just 1 homer over his 51 innings, he sports a microscopic 2.8% HR/FB rate. In a park that inflates home runs by 8%, the rate in which his fly balls go over the fence is going to rise. Given what appears to be an extremely fortunate performance so far, it comes as no surprise that if he qualified, he would easily rank first in ERA-xFIP differential among all starting pitchers. His 4.56 xFIP sits nearly three full runs above his ERA.
As much as I stress the importance of looking beyond ERA and at the component skills instead, it’s still hard to actually put that into practice. I see that first hand when reading the comments section of my AL SP tier rankings each month! But the majority of the time, a pitcher’s ERA will eventually match his skills, which are summarized by the ERA estimators like SIERA and xFIP. This of course assumes that the pitcher maintains his peripherals to begin with, but that’s an article for another day.
Shamelessly cherry picking an example, Jeff Locke was one of the luckiest pitchers in baseball all season long. While his skills did dip in August, they were never very good to begin with. After posting a 2.19 ERA over his first 20 starts, the implosion finally came starting with his July 31st start, as he then posted an 8.71 ERA over his last 6 starts. This ultimately led to his demotion yesterday.
While I cannot speak for every fantasy owner of Locke’s, I am fairly confident that despite his large season long ERA/SIERA discrepancy, he continued to be started each week without question. The mistake being made is thinking of ERA as some sort of skill metric or a badge of what that pitcher is. It’s not. It’s simply the result of multiple underlying skills that get thrown into a blender and ultimately yield the various surface stats like ERA and WHIP.
So the lesson is that while I understand using Locke as a cherry picked example proves little, betting on Cosart, and any pitcher whose ERA is significantly below what his peripherals suggest, is a losing proposition. The outlook on Cosart long-term is entirely different, however, so do not interpret this piece as a pessimistic vote for his future. My opinion is far from that and there are several things I really like about him that does make him intriguing. But for the rest of this season, I wold find it difficult to start him even in AL-Only leagues.