May vs. Pelfrey The Actual Data and The Law of Unintended Consequences

[Otto von Ballpark]

 

The injury part is fair.  The getting pulled early part I don't buy nearly as much.  I remember a Nat. League start where he was pulled for a pinch hitter.  Other than that, he was pulled when his control started to leave him and he was starting to get hit more.

Fine, let's say May had exactly the same leash as Pelfrey.  Still, on average, tossing aside that one injury shortened start, Pelfrey lasted less than two-thirds of an OUT longer than May.  Not two-thirds of an inning -- less than two-thirds of one OUT.  20% of one inning.

[Jham]

 

Fine, let's say May had exactly the same leash as Pelfrey.  Still, on average, tossing aside that one injury shortened start, Pelfrey lasted less than two-thirds of an OUT longer than May.  Not two-thirds of an inning -- less than two-thirds of one OUT.  20% of one inning.

 

If you want to total the innings and average them, great.  I guess it depends on whether you like shorter possibly better starts, or longer possibly less effective starts.  Pelfrey and May play in front of the same defense.  Why are the results so different?  Defense?  I'll take Pelfrey's stretch over May's.  This may change.  Pelfrey seems to have lost the splitter which was a difference maker.  All I'm saying is that Pelfrey's value is his ability to limit HR (shows up in FIP) generate double plays (doesn't show up in FIP) and go 9/20 starts 7 innings or more.  May's biggest knock, like it or not, is that he hasn't gone deep in games, and that his walk rate and HR rate are far FAR below anything he's done in his career.

 

[Mike Sixel]

 

If you want to total the innings and average them, great.  I guess it depends on whether you like shorter possibly better starts, or longer possibly less effective starts.  Pelfrey and May play in front of the same defense.  Why are the results so different?  Defense?  I'll take Pelfrey's stretch over May's.  This may change.  Pelfrey seems to have lost the splitter which was a difference maker.  All I'm saying is that Pelfrey's value is his ability to limit HR (shows up in FIP) generate double plays (doesn't show up in FIP) and go 9/20 starts 7 innings or more.  May's biggest knock, like it or not, is that he hasn't gone deep in games, and that his walk rate and HR rate are far FAR below anything he's done in his career.

 

9 of 20? That's not all that good.....

 

It is better to go 6 innings and get pulled, before coming into the 7th and being bad. Just look at Pelfrey's stats since 2010, and tell me he is a guy you WANT starting over May or Berrios.

[Mike Sixel]

Wait, boyer is in, and not May or Jepsen?

 

What?

[drjim]

 

You did notice that I acknowledged that fact in the first paragraph? 

While Pelf's been a pleasant surprise at times... there's no denying that he's also been the model of inconsistency, especially in his last 9 starts, and there's nothing to chuckle about with his total net results.

 

And honestly ask yourself this question, who would you rather have start in a key game in the stretch run, or say, Game 3 in the ALDS?

 

I also said everything you said is pretty much true, so that should probably answer the question.

 

That said, two basic stats about May as a starter that need to be continually mentioned (and not just excused away) - Leaving aside Nolasco who is out for the year, May has averaged the fewest innings per start and has given up the most runs per inning as a starter. That is about as basic as it gets, so I don't exactly think he is a slam dunk guy in the rotation.

 

I do think he is likely to be one of best (certainly no worse than 4th) going forward, but it wasn't an atrocity at the time.

[Otto von Ballpark]

 

That said, two basic stats about May as a starter that need to be continually mentioned (and not just excused away) - Leaving aside Nolasco who is out for the year, May has averaged the fewest innings per start and has given up the most runs per inning as a starter.

When the decision was made:

 

May: 5.4 IP/GS, 4.59 runs per 9 innings

Pelfrey: 5.7 IP/GS, 4.14 runs per 9 innings

 

Less than a third of a inning difference, less than a half a run difference per nine innings.  I doubt these numbers were involved with the decision, except as an after-the-fact justification.

[drjim]

 

When the decision was made:

 

May: 5.4 IP/GS, 4.59 runs per 9 innings

Pelfrey: 5.7 IP/GS, 4.14 runs per 9 innings

 

Less than a third of a inning difference, less than a half a run difference per nine innings.  I doubt these numbers were involved with the decision, except as an after-the-fact justification.

 

I never even heard them quoted so probably weren't even a justification. This was my own sleuthing!

 

I would prefer May myself for all the reasons that have been cited, but it's not like it was an injustice or even bad strategy if May is better out of the pen.

[Otto von Ballpark]

 

I would prefer May myself for all the reasons that have been cited, but it's not like it was an injustice or even bad strategy if May is better out of the pen.

Agreed.  But they needed to be more aggressive with May in the bullpen to really justify it, and/or more willing to move May back to the rotation when needed.  They've failed on both counts so far, and it's cost us games in the former.  Hopefully Duffey steps up and saves them from the latter.

[drjim]

 

Agreed.  But they needed to be more aggressive with May in the bullpen to really justify it, and/or more willing to move May back to the rotation when needed.  They've failed on both counts so far, and it's cost us games in the former.  Hopefully Duffey steps up and saves them from the latter.

 

It was indeed a little slow to get rolling but May is clearly being used as the 8th inning guy right now, even if the recent run of futility has masked it.

[Otto von Ballpark]

 

It was indeed a little slow to get rolling but May is clearly being used as the 8th inning guy right now, even if the recent run of futility has masked it.

Yes, I have noticed.  But prior to this recent use, I think they squandered a couple games without using May.  (And "8th inning guy" may still be holding him back a bit, no reason he can't be a multi-inning reliever when the situation warrants.  I'd rather pitch May for a couple innings in a close game and get by the next game without him, than need another reliever to help in addition to May in every close game.)

[drjim]

 

Yes, I have noticed.  But prior to this recent use, I think they squandered a couple games without using May.  (And "8th inning guy" may still be holding him back a bit, no reason he can't be a multi-inning reliever when the situation warrants.  I'd rather pitch May for a couple innings in a close game and get by the next game without him, than need another reliever to help in addition to May in every close game.)

 

I think that is probably the next move.

[Otto von Ballpark]

I think that is probably the next move.

Seems backwards? To move to 1 inning relief first, then multi inning relief?

 

Sometimes I think this team gets caught up in ideals too much. Their ideal bullpen is a bunch of good 1 inning guys. And we waste time trying to fit that ideal first than using the resources we have optimally. (Also see the SS situation)

[drjim]

 

Seems backwards? To move to 1 inning relief first, then multi inning relief?

Sometimes I think this team gets caught up in ideals too much. Their ideal bullpen is a bunch of good 1 inning guys. And we waste time trying to fit that ideal first than using the resources we have optimally. (Also see the SS situation)

 

I think it will be more out of desperation than anything.

[Oxtung]

I think the biggest reason to put May back into the rotation is his K-rate. At ~8 his K/9 is above average and over the last 3 years pitchers with an above average k-rate have had above average ERAs 75% of the time. Those with below average K-rates have above average ERAs just 39% of the time. Those that are well below average, like Pelfrey, are above average rarely.

 

 

*I understand I'm comparing FIP and xFIP, FIP seems more appropriate for Pelfrey who has posted a FIP below his xFIP virtually every year of his career. I used xFIP for May because his profile is that of a pitcher who will give up an avg. to above avg. HR total.

**Additionally, I have some concerns about FIP and xFIP in general, as they start with a somewhat flawed premise that pitchers can't control contact. They eliminate 1 huge variant (defense), but don't account for GB/FB, induced double plays, etc that don't result in home runs. It's a far from perfect stat.

 

 

 

Then you should use the ERA estimator (and really I think everybody should be anyways) SIERA which accounts for a pitchers batted ball profile and adjusts for league and ball park. Though, SIERA says Pelfrey is still a significantly worse pitcher than May.

 

Also May has had very good HR rates in the minors as well. Unlike his walk rates there is nothing to suggest he should regress significantly.

[Jham]

 

I think the biggest reason to put May back into the rotation is his K-rate. At ~8 his K/9 is above average and over the last 3 years pitchers with an above average k-rate have had above average ERAs 75% of the time. Those with below average K-rates have above average ERAs just 39% of the time. Those that are well below average, like Pelfrey, are above average rarely.


Then you should use the ERA estimator (and really I think everybody should be anyways) SIERA which accounts for a pitchers batted ball profile and adjusts for league and ball park. Though, SIERA says Pelfrey is still a significantly worse pitcher than May.

Also May has had very good HR rates in the minors as well. Unlike his walk rates there is nothing to suggest he should regress significantly.

 

I like the notion of trying to factor for more variables.  SIERA is still based on general premises that may or may not be true in a pitcher's specific case (Fly ball pitchers limit HR, strike out pitchers have a higher GB/DP percentage).  

I also REALLY don't understand the idea that ERA is a flawed statistic, but SIERA and FIP are both attempting to predict future ERA...  I'd rather see FIP predict future FIP and SIERA future SIERA.  Why try to predict something we don't feel like using anyway?  I also don't understand why taking defense out of the equation is so important when comparing teammates who play in front of the same 8 defenders...  (Pace of play, ability to pound the strike zone, and your own ability to field the 1 spot may account for some variance, but those are pitcher controlled anyway)