Of pitchers with 10 IP or more in the Nats bullpen, Tyler Clippard has the best ERA (1.72) and Tyler Walker has the second worst (3.94). But who is really the worse pitcher?
ERA does a poor job measuring a pitcher's performance. That is because ERA is computed from ER which is determined by many things beyond a pitcher's control such as his team's defense and the stadium he plays in. tERA, however, is a function of events (mostly) under a pitcher's control such as K, BB, and GB%. In this category Clippard has a 2.36 and Walker a 3.48-the two best marks in the bullpen. Things have gotten much closer, but we are not finished.
While tERA is, generally speaking, one of the best single pitching quantifiers around, in this case it is incomplete. If we assume that Clippard can sustain his current tERA, then we're also essentially assuming he can maintain his incredibly low 3.8 HR/FB%, a metric that has been argued to be mostly beyond a pitcher's control, and, more immediately, a rate well below Clippard's career line, amongst other assumptions. When we consider xFIP, which considers HR/FB% as being beyond a pitcher's control, we see that Walker comes out as the superior pitcher with an xFIP of 3.79 versus Clippard's xFIP of 3.96.
Also, because of the small sample size of innings, it is hard to truly tell how well either pitcher is playing. But we can be certain that ERA is misleading. Some other things to consider are Clippard's dangerously low GB% (29.2) and dangerously high FB% (58.4). Will his propensity to give up fly balls turn into a propensity to give up home runs? Then there's Clippard's extraordinarily high LOB% (89.3). Is this something Clippard can control or not? If so, should it be reflected in statistics claiming to measure his worth? In the above categories Walker has numbers much closer to league averages, another reason for believing that he can maintain his current level of success while Clippard cannot. There's also Walker's K/BB rate of 3.86 which is superior to Clippard's (2.71).
The lesson to take away from this is to not let discussions about pitchers end at ERA. If we did, we'd be doing a disservice to all those who want to better understand baseball. Worse, we'd be arguing that Miguel Batista (tERA 5.79, xFIP 5.89, K/BB 0.81) is pitching about as well as Walker-something that should be plain indefensible.