FLUKE OR NOT?--PART 3

by Matt Swartz on Jun 6, 2008 2:40 AM EDT

I'll finish off the final part, part 3 of this series, by studying a few relief pitchers-- Brad Lidge, Chad Durbin, and J.C. Romero. Unsurprisingly, their performances are all pretty flukey. The Phillies bullpen has been great this year, and these three guys are a big part of the reason why. In fact, the bullpen has been a main contributor to the team's success. My best guess is that if the Phillies are going to maintain the pace they have now, they are going to need to get some improvement from the hitters and the starting rotation, because the bullpen is bound to regress to the mean. That's not to say that the bullpen won't still be pretty good, but it won't be as reliable as it has been.

Brad Lidge:

Actual: 0.72 ERA

Projected (ZIPS): 3.88 ERA

His ERA is 0.72, so clearly he isn’t going to maintain this, but he certainly is having a good year. However, his K rate is lower than his career rate (10.80 opposed to 12.52) and his BB rate is higher (4.68 above 3.83), but his he has yet to give up a homerun all year. In his career, he has given up 10.4% homeruns/flyball and this year, that would imply that he should give up two homeruns, which would generally lead to about three more runs, and have an ERA closer to 1.80 instead of 0.72, and his BABIP is only at .229 instead of his career .318. That’s going to turn around too, as his groundball rate is actually slightly higher (which correlates to higher BABIP).

My guess: FLUKE. He’ll still have about a 3.30 ERA from here out, good to finish around 2.40 or so. He’ll still win a few more closer games for the Phillies, but he’ll look less like superman in the process.

Chad Durbin:

Actual: 1.80 ERA

Projected (ZIPS): 4.43 ERA

His walk and strikeout rates seem pretty similar to his career rates, but his BABIP is way down at .258 this year, despite having similar groundball/flyball/line drive rates. His homeruns/flyball is down at 2.7%. That’s not going to stick either. This is not going to stick.

My guess: BIG FLUKE. Don’t canonize the man. He’ll show his true colors soon.

J.C. Romero:

Actual: 1.66 ERA

Projected (ZIPS): 4.17 ERA

His walk rate is still way too high, sitting around 5.82 now. His strikeout rate is up at 7.89, certainly higher than it’s been, but not really that much higher than usual. His groundball rate is still around 60%, very similar to last year. Despite only giving up four of his own runs, he’s apparently given up six of other people’s runs, letting six of 16 inherited runners score. I think his ERA is low because he’s entering innings with one or two outs already.

My guess: FLUKE. His ERA will still belie a mediocre performance, but some of the runs he lets in will eventually be his own.

Tweet Comment 6 comments | 0 recs |

Do you like this story?

Comments

I agree that the bullpen is overperforming, but also that it’s a solid group—the best or close to the best they’ve had in 25 years or so.

On Lidge, I remember seeing the projections for him at the start of the season and thinking they were all too high: if he was healthy, I thought he’d be effective. Not 1.00 ERA-type effective, but all-star closer-effective. Romero and especially Durbin are bound for a comedown, but hopefully improvement elsewhere-Ryan Howard, I’m looking in your direction-will make up for that.

by dajafi on Jun 6, 2008 1:15 PM EDT reply actions

On my count, Romero’s entered an inning in with one or two outs 10 times, and with zero outs 16 times.

by taco pal on Jun 6, 2008 1:34 PM EDT reply actions

Romero

One thing about Romero that I think most statistical projections miss is that he’s uniquely suited to getting out of jams (his own or others), because of his ability to strike hitters out and induce ground balls. So while his walk rate is high, he’s probably less likely than most to be hurt by it because the next batter is likely to strike out or GIDP.

Also, his ability to wriggle off the hook is probably responsible for his usage pattern. More than any other reliever he’s brought in when there’s trouble brewing. This means that he’s likely to come in with runners in scoring position, possibly with fewer than two outs, etc. This will result in him letting inherited runners score more often than you might otherwise expect.

Mind you, I don’t think Romero is as good as his ERA, but he’s a really good reliever.

by phatj on Jun 6, 2008 1:42 PM EDT reply actions

True, his usage is important. He’s a high OBP, low SLG kind of guy. Those types of guys are better used with 1 or 2 outs in an inning. I read that the reason Tom Glavine perpetually has a lower ERA than his DIPS would indicate (career FIP 3.93, career ERA 3.52) is due to his approach. He nibbles with runners on base late in the inning and pounds the strike zone with bases open early in the inning. The reason is that the damage of a walk relative to an extra base hit is larger earlier in the inning. Romero is definitely being used quite effectively. However, I still think he’s going to let in more 1-out and 2-out runs than he has been.

by Matt Swartz on Jun 6, 2008 4:00 PM EDT up reply actions

I’m confused, so ERA define whether or not a guys a fluke?

by jemagee on Jun 6, 2008 2:18 PM EDT reply actions

Yes and no. On one hand, the main goal of pitching is run prevention. The value of the outcome comes from preventing runs, and therefore ERA is my summary statistic of the outcome of pitcher performance—whether as a result of luck or not.

When it comes to mid-inning relief, that number does a poorer job and something like WPA or WXRL can be helpful, but for guys like Lidge, the value of his performance comes from a low ERA for the most part.

Whether or not a pitcher is a fluke is basically me saying whether or not I think that pitcher will prevent runs at the rate that they have been.

by Matt Swartz on Jun 6, 2008 3:56 PM EDT up reply actions