Starting Pitchers ERA-xFIP Differential Update

It has been nearly a month since I first took a look at the differential between starting pitchers’ ERA and xFIP. With two months now in the books, it is time for an update. From message board posts and various articles, it seems that many fantasy owners wait until June to really analyze their team and look for trading opportunities. Determining which pitchers have been the luckiest and which have been the most unfortunate is a good exercise to undertake to identity potential trade targets or sell high candidates.

Funny, Jair Jurrjens was one of just three pitchers I singled out as original members of The Overvalued Club this pre-season, and I was actually just called out on it two days ago. Jurrjens’s strikeout rate is at the lowest of his career, as is his SwStk%, and his fastball velocity is down 1.5 miles per hour. Sounds like a recipe for a disastrous season, huh? Well, he has offset that with an elite walk rate, despite the fact he is throwing just about the same percentage of first strikes as usual, and has actually thrown the fewest pitches inside the strike zone in his career. So that looks quite unsustainable. The GB% spike is good to see, though, and may continue as he has shown that ability in the past, back in 2008. So not only will Jurrjens’ luck run out at some point, but his walk rate is likely to jump, combining for a double whammy. Bold prediction: Jurrjens posts an ERA above 4.00 for the rest of the season. Owners should be trying to trade him as aggressively as humanely possible.

Hmmm, Josh Beckett is pitching almost identically to last year from a skills perspective, with the exception of a slight drop in his GB%. Yet, his ERA has dropped four full runs so far. His velocity is down again, for a second straight year, and now at a career low. However, it has not affected his SwStk% or K/9 yet, so maybe it is nothing to worry about. I have always been a big Beckett fan and would expect around a mid-3.00 ERA going forward. Of course, like any pitchers on the lucky list, it couldn’t hurt to at least shop him around to see if you could pick up a top bat in return.

I have written about Jeremy Hellickson several times, including making an official bold prediction during the pre-season that he would be the most valuable Rays pitcher. Of course, I made that prediction with the expectation that it would be a skills-fueled performance, not luck-fueled. Despite the 2.80 ERA, he has actually been a disappointment, as his walk and strikeout rates are both worse than I had forecasted. The good news is that his SwStk% suggests a much better K/9 and his F-Strike% hints at better control moving forward. As a result, I think his peripherals will improve, narrowing the gap between his ERA and xFIP before that ERA really has a chance to spike to dramatically. Therefore, Hellickson owners should probably hold.

Man, it’s a shame that Josh Johnson simply cannot stay healthy. Ground balls? Check. Strikeouts? Check. Excellent control? Check. A top talent for sure.

Alexi Ogando pitched yesterday, but his start is not reflected in the above table. His xFIP definitely improved, but I am not sure whether the differential narrowed. Obviously, his owners are loving this start as he likely came rather cheaply and simply refuses to experience the regression that many have been assuming will come. He is a tough one when it comes to the trade market because he still may not garner much in return, especially in single-season leagues. In keeper leagues, he may be a very attractive target for dumping teams. Keeper league owners going for it this year may be able to get a lot for him and I would highly suggest making the move.

After another clunker, Ryan Dempster‘s ERA and xFIP are one the rise once again. However, he is pitching almost exactly the way he pitched last year, even posting a nearly identical xFIP-. His fastball velocity is down a tick and the pitch has gotten crushed according to his pitch type values. This also seems to be hurting his slider’s effectiveness, as right now the pitch is worth less than 1.0 runs above average per 100 pitches for the first time since 2004. His SwStk% and Contact% are way off from previous seasons, which makes it curious that his K/9 has not dipped. That likely has to do with my gripe of using outs as the denominator rather than plate appearances. Looking at K/PA, his rate has indeed dropped from 22.7% to 20.3%. Though he clearly has been unlucky as he is not truly deserving of his current 6.32 ERA, it seems like something clearly is not completely right here, but unfortunately I don’t know what that is. Buy low if it is low enough, but obviously do not value him as you did in the pre-season.

I could have told you that after allowing at least four runs in each of his last four starts, Bronson Arroyo was guaranteed to get good results now that he was facing my pitcher. Rants about not getting any run support aside, as unlucky as Arroyo may be, I still have little interest, even in NL-Only leagues.

Fausto Carmona‘s primary problem is his ridiculous 56.6% LOB%. However, his extreme ground ball ways makes him like a mini Chien-Ming Wang, but with greater strikeout potential. His skill set is unlikely to generate significant value in mixed leagues, but he looks like a nice target in AL-Only leagues.

Chris Narveson‘s ERA is somewhat perplexing as he has suffered from a bit of bad luck on the BABIP and LOB% side, but not significantly, and his HR/FB ratio is actually below league average. He was a popular add at the beginning of the season when he got off to a fast start, but he has probably been dropped by now in all the shallower leagues he was added in. His SwStk% suggests even further K/9 upside, so better luck may even earn him some mixed league value. If nothing else, he makes for an excellent NL-Only target.

Not too interested in Brian Duensing as his GB% is down and his increase in strikeout rate has been offset by worse control. He lacks the skills upside and strikeout potential to really be a good AL-Only target.




Print This Post

Mike Podhorzer produces player projections using his own forecasting system and is the author of the eBook Projecting X: How to Forecast Baseball Player Performance, which teaches you how to project players yourself. His projections helped him win the inaugural 2013 Tout Wars mixed draft league. He also sells beautiful photos through his online gallery, Pod's Pics. Follow Mike on Twitter @MikePodhorzer and contact him via email.


15 Responses to “Starting Pitchers ERA-xFIP Differential Update”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. odditie says:

    What about Brandon Morrow?

    ERA: 5.11
    FIP: 2.08
    xFIP: 3.11
    SIERA: 2.95

    Vote -1 Vote +1

    • Mike Podhorzer says:

      Missed the minimum innings qualifier. However, he makes for a great buy low guy.

      Vote -1 Vote +1

      • odditie says:

        I figured that was what held him off the list, but I’ve been holding out hope on him in my league. Every bad start makes me wish I didn’t know those other metrics almost.

        Vote -1 Vote +1

  2. Travis says:

    Someone who just missed the cut as a negative regression candidate, Charlie Morton, is throwing more offspeed and curveballs to LHB tonight against Philly. Throwing a few sliders, too (might be cutters, can anyone confirm the pitch f/x movement? I’m not great with the movement aspect of pitch f/x). If those sliders are cutters…

    Vote -1 Vote +1

  3. Antonio Bananas says:

    So, FIP says that guys with ridiculously low ERAs won’t sustain that and guys with ridiculously and uncharacteristically high ERAs will pitch better? Man, and all this time I’ve just been looking at guys like Ubaldo Jimenez last year and saying “no way he does that all year”. I should have looked at this FIP thing instead. So useful.

    -8 Vote -1 Vote +1

  4. Yirmiyahu says:

    Mike P,

    I hope you’ve seen this: http://www.hardballtimes.com/main/blog_article/why-you-cant-subtract-fip-from-era

    With Josh Johnson, for instance, his xFIP (3.14), FIP (2.56), and ERA (1.64) are all perfectly in line with each other.

    Vote -1 Vote +1

    • Elias says:

      I think that HBT post overstates the problem. Some of the reason for the “flatness” of FIP and xFIP is exactly the reason why subtraction makes sense: ERA noise. If ERA is a noisy estimate of talent and xFIP is a precise estimate of talent, then you’d expect to see the flatter xFIP trend in the HBT figure. Some of it could also be due to the linear model of run scoring, but the HBT doesn’t make that case.

      (The “problem” with the figure in that model is that the x-axis should be some measure of true talent, and not the noisy ERA measure. Measurement error in an “X” variable does very bad things in regression land.)

      Vote -1 Vote +1

      • Yirmiyahu says:

        I understand that, and his methodology had some obvious flaws. But we’re not trying to compare FIP or xFIP against true talent.

        By looking at ERA-xFIP differential, we’ve already decided to use the “noisy ERA measurement”. We’re not trying to measure xFIP against true talent. The question is “What is the expected ERA for someone with x.xx xFIP?” If there is good data showing that someone with a 3.00 xFIP usually has a 1.50 ERA, it doesn’t matter to me whether it’s because of the flaws with ERA, or because of true talent.

        Vote -1 Vote +1

      • Elias says:

        Not to belabor the point, but the HBT piece does NOT provide evidence that someone with a 3.00 xFIP usually has a 1.50 ERA, rather they show that someone with a 1.50 ERA usually has a 3.00 xFIP. This is perhaps a clearer way to illustrate my technical point about measurement error. The majority of players who post 1.50 ERAs are NOT true talent 1.50 ERA pitchers. They are getting lucky, and the fact that they tend to post much higher xFIPs is a count in favor of looking at the difference, not a strike against.

        Vote -1 Vote +1

  5. William Smith says:

    Thank you for the very good article!

    RE: Dempster – “Though he clearly has been unlucky as he is not truly deserving of his current 6.32 ERA, it seems like something clearly is not completely right here,”

    I think he might deserve it. Right now, Dempster has an unnaturally high HR/FB% (16.9%) and an SUPER low LOB% (64.4%)

    All 5 of your high-ERA guys either have extremely low LOB%, extremely high HR/FB%, or both.

    Seems to me, a large part of the differential between xFIP and ERA, we might be able to explain by these guys deviating so much from league averages in LOB% and HR/FB%. Its interesting that FIP appears to account for the difference in Homerun Rates, while xFIP doesn’t.

    6/5/2011
    Name LOB% HR/FB ERA FIP xFIP
    Ryan Dempster 64.40% 16.90% 6.32 4.63 3.53
    Bronson Arroyo 71.10% 14.10% 5.37 5.14 4.14
    Fausto Carmona 57.70% 14.30% 5.33 4.42 3.76
    Chris Narveson 67.50% 8.60% 4.68 3.26 3.28
    Brian Duensing 65.00% 8.50% 5.37 3.96 4.02
    2011 LEAGUE AVG 73.2% 9.1%

    Many of these guys’ metrics for LOB% and HR/FB% are >1STDV, and in some cases >2STDV. They’re near the tails of the distribution & to me it seems like xFIP flattens their effect when they’re on the extreme.

    Would very be interesting to see if there’s a relationship between deviation from these metrics and xFIP-ERA differential…..

    Vote -1 Vote +1

    • Mike Podhorzer says:

      Well sure, if you assume the high HR/FB ratio and resulting low LOB% is not bad luck, but bad pitching, then Dempster is deserving of a high ERA. But we typically treat those metrics are primarily luck based, unless the pitcher has shown a consistent tendency in past seasons to over or under perform the league average.

      Vote -1 Vote +1

  6. Antonio Bananas says:

    You should do another article of guys who’s ERA and FIP are closest, meaning they should stay about the same the rest of the year and see if that holds true.

    Vote -1 Vote +1

  7. JBL says:

    Here is where I will call you out and add just a hint of perspective. While the numbers can and will always be helpful in evaluating players, it helps to actually see the Pitcher pitch.

    Case in point: Josh Beckett. It matters not that his FB velocity is down. What matters is that FB, which has been as flat as a board for the past two seasons, finally has some deceptive movement on it again. That and the fact that he has the most stingy defense in baseball playing behind him. So FIP is somewhat neutralized, because as long as the BoSox stay healthy they are going to help their pitchers win games.

    As for Jurrjens, I agree there will be some correction in his numbers coming. But it’s also important to consider two non-numerically quantifiable factors – he’s healthy for the first time in almost two years and he learned, adapted and has been effectively using Venters cutter which apparently looks just like his FB in his delivery.

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>