Should Bobby Parnell be Mets’ 8th-inning reliever?

Coming into the 2011 season, the Mets were counting on Bobby Parnell to be their 8th inning reliever, the main bridge to closer Francisco Rodriguez. But Parnell was terrible and soon found his way on the DL. Even after he was healthy, Parnell spent extra time in Buffalo trying to get back on track. Since he returned at the end of May, Parnell has been lights out, to the point where perhaps it’s best if he’s again the club’s first option for the eighth inning.

But would it be a good idea to have Parnell as your main setup man? In April, he had a 13.50 K/9 and an 80.4 strand rate. If you just had those two numbers to look at, you would conclude that he had been very effective when he was anything but that. So what makes Parnell effective, or in the case of April ineffective? Let’s look at his numbers.

None of these are a huge sample, so it’s okay if you want to take these numbers with a grain of salt. But one thing that jumps out is that it almost does not matter what Parnell’s strikeout rate is, which is virtually unheard of for a pitcher. As noted earlier, he had a tremendous K/9 in April this year and was terrible.

Another thing that stands out is that Parnell has very high BABIPs. Even his numbers as a reliever in 2009 are a bit misleading. If you remember that season, Parnell had great success as a reliever and then had an audition as a starter before ending the year back in the pen. Parnell had a .345 BABIP as a reliever before moving to the rotation. When he returned to the pen, he had an .071 BABIP which lowered his overall numbers.

Parnell lights up radar guns and racks up strikeouts, which seems like exactly what you want from a late-inning reliever. But he’s also the most hittable reliever among guys who throw gas. Using the FanGraphs leaderboards, for the past two years, Parnell has the fifth-highest average fastball velocity. Here are all the guys who average 96 and above and their corresponding BABIPs:

There were 261 relievers in this sample and Parnell’s BABIP was the fifth-worst mark in the bunch.

So, we expect Parnell to rack up an impressive strikeout rate and we expect him to have a high BABIP. Therefore Parnell’s success depends more on his walks and home runs allowed than anything else. When he struggled as a SP in 2009 and then again early in 2011, he had trouble with the gopher ball. Those same two spans he had a BB/9 greater than 5.00.

Mets fans already have a closer who walks the tightrope by allowing lots of baserunners in Rodriguez. Do they want another tightrope walker for their primary setup man? For his career in the majors, Parnell has a 1.56 WHIP. When he limits his walks, like he did last year, he has a 1.40 WHIP, which was in the bottom third of relievers with at least 30 IP.

But much like with Rodriguez, the total package seemingly works. Even though on first glance he appears very fortunate, last year he had an xFIP of 2.54 and this year it’s 3.01. His ultra-high BABIPs apparently cancel out his elevated strand rates and his low HR marks, given his K-rate. Those huge strikeout numbers pull him through, after all.

It would not be crazy to reinstall Parnell back as the primary setup man, perhaps in a time-share with Jason Isringhausen, who should not pitch on back-to-back days. Eight times this year, Isringhausen has pitched with no rest and he has an .861 OPS allowed in those games. When he has a day of rest between appearances, his OPS is .454, with two days rest it’s .634 and it’s .298 when Izzy’s had three days rest.

Still, perhaps the Mets can rig things so that Parnell always comes in to start an inning. This year he has come in with eight runners on base and allowed six runners to score. By contrast, Isringhausen has come in with 11 runners on base and allowed only two to score.