The Yankees Bet on Brandon McCarthy and xFIP

The Yankees just pulled a rare feat by trading Vidal Nuno to the Diamondbacks for Brandon McCarthy. Only once in the last five years has a team traded for a pitcher whose results were so out of whack with their process and peripherals. Of course, that was when the Dodgers traded a player to be named later to the Phillies for Joe Blanton in 2012, but the Yankees have a few reasons to believe that this will turn out better for them than that trade did for the Dodgers.

ERA does not tell the full story of Brandon McCarthy‘s season so far. Look across his line, and you see career-best strikeout (20%) and ground-ball rates (55.3%) paired with his customary excellent command… and then you see that he’s giving up twice as many home runs on fly balls as he has his whole career.

Let’s focus first on the stats by themselves, without context. We can look at his xFIP for short-hand on his walk, strikeout and fly-ball rates, and we can see a career-best number there. And we can see the big gap between his ERA and xFIP. And we can look for the 20 other qualified starting pitchers that have had the biggest such gaps in the first half. And we can create this list:

Season Name K/9 BB/9 HR/9 BABIP LOB% GB% HR/FB ERA xFIP xFIP-ERA
2012 Tim Lincecum 9.7 4.7 1.0 0.333 59.2% 43.5% 12.4% 6.42 3.88 -2.54
2009 Ricky Nolasco 8.9 2.1 1.2 0.341 58.1% 36.3% 10.7% 5.76 3.46 -2.30
2012 Jake Arrieta 7.9 2.8 1.2 0.324 58.2% 42.9% 12.9% 6.13 3.84 -2.29
2014 Brandon McCarthy 7.6 1.6 1.2 0.345 66.7% 55.3% 20.0% 5.01 2.89 -2.12
2013 Wade Davis 8.3 3.9 1.2 0.381 66.3% 38.3% 13.5% 5.89 3.99 -1.90
2013 Rick Porcello 7.4 1.8 1.0 0.323 65.4% 57.1% 15.9% 4.90 3.02 -1.88
2013 Joe Blanton 7.5 2.1 1.8 0.343 70.6% 44.2% 18.1% 5.53 3.74 -1.79
2014 Ricky Nolasco 6.3 2.4 1.4 0.362 68.7% 40.8% 12.2% 5.90 4.15 -1.75
2013 Edinson Volquez 7.7 4.2 0.7 0.342 63.3% 48.3% 8.8% 5.74 4.05 -1.69
2010 Kevin Millwood 7.0 2.9 1.7 0.337 67.6% 38.6% 14.8% 5.77 4.11 -1.66
2011 Roberto Hernandez 5.4 2.9 1.3 0.288 59.4% 57.9% 15.6% 5.78 4.14 -1.64
2010 James Shields 8.5 2.0 1.5 0.330 67.7% 40.5% 14.3% 4.92 3.32 -1.60
2010 Scott Kazmir 5.9 4.8 1.7 0.288 63.3% 40.1% 13.0% 6.92 5.33 -1.59
2011 Chris Volstad 5.9 2.7 1.4 0.309 66.4% 51.2% 16.5% 5.40 3.82 -1.58
2012 Joe Blanton 7.8 1.3 1.6 0.305 66.1% 42.5% 16.9% 4.92 3.43 -1.49
2010 Nick Blackburn 3.2 2.5 1.8 0.326 66.2% 48.0% 14.8% 6.40 4.92 -1.48
2009 Cole Hamels 7.8 1.7 1.4 0.343 71.3% 39.4% 13.5% 4.87 3.43 -1.44
2009 Scott Baker 7.3 1.9 1.6 0.282 62.8% 32.8% 12.6% 5.42 3.98 -1.44
2012 Adam Wainwright 8.6 2.5 0.9 0.333 67.7% 51.8% 13.9% 4.56 3.12 -1.44
2011 Ryan Dempster 8.3 3.2 1.1 0.326 68.3% 46.0% 12.8% 5.01 3.58 -1.43
  Average 7.3 2.7 1.3 0.328 65.2% 44.8% 14.2% 5.56 3.81 -1.75

Obviously, it takes a confluence of track record and bad luck to get on this list. If you haven’t shown some promise, you’re not going to continue getting chances. But these pitchers also had terrible numbers in the parts of the game where they haven’t been shown to have great control over results. This group’s strand rate (65.2%) and home run per fly ball rate (14.2%) in particular, are well above the league’s number (generally 70% and 10% in any given year). And not in a big enough sample to believe they’ve earned those numbers.

So what did this group do in the second halves after their disastrous starts?

Season Name K/9 BB/9 HR/9 BABIP LOB% GB% HR/FB ERA xFIP xFIP-ERA
2012 Tim Lincecum 8.7 4.0 1.2 0.281 79.1% 48.6% 17.4% 3.83 3.75 -0.08
2009 Ricky Nolasco 10.0 2.2 1.1 0.290 64.4% 40.6% 11.2% 4.39 3.00 -1.39
2013 Wade Davis 5.9 3.8 0.6 0.361 69.0% 42.2% 5.6% 4.99 4.72 -0.27
2013 Rick Porcello 7.1 2.8 0.8 0.310 74.6% 52.4% 12.1% 3.82 3.49 -0.33
2013 Edinson Volquez 7.2 3.9 1.5 0.294 66.7% 47.4% 17.9% 5.73 4.08 -1.65
2010 Kevin Millwood 5.2 3.3 1.1 0.291 74.8% 35.4% 8.1% 4.23 4.92 0.69
2011 Roberto Hernandez 5.0 2.9 0.8 0.296 65.1% 50.9% 9.6% 4.59 4.21 -0.38
2010 James Shields 8.1 2.6 1.6 0.358 68.9% 41.7% 13.2% 5.59 3.82 -1.77
2010 Scott Kazmir 5.2 4.7 1.3 0.270 76.4% 37.5% 9.4% 4.37 5.55 1.18
2011 Chris Volstad 7.1 2.6 1.0 0.313 73.1% 54.5% 13.7% 4.04 3.35 -0.69
2012 Joe Blanton 7.9 2.0 1.0 0.314 70.1% 47.7% 13.0% 4.35 3.26 -1.09
2010 Nick Blackburn 4.7 1.9 0.7 0.258 72.1% 57.1% 9.6% 3.54 3.76 0.22
2009 Cole Hamels 7.8 2.4 0.9 0.289 73.1% 41.6% 7.9% 3.76 3.84 0.08
2009 Scott Baker 7.3 2.5 0.9 0.271 77.9% 34.0% 6.9% 3.28 4.32 1.04
2012 Adam Wainwright 8.1 2.2 0.5 0.296 67.9% 49.6% 6.3% 3.28 3.35 0.07
2011 Ryan Dempster 8.6 3.1 1.3 0.313 69.1% 44.7% 14.1% 4.76 3.81 -0.95
  Average 7.1 2.9 1.0 0.300 71.4% 45.4% 11.0% 4.28 3.95 -0.33

Better. Their collective ERA dropped a full run, as did the gap between their xFIP and ERA. They gave up fewer home runs, too, on a more normal home run per fly ball rate. Their strand rate regressed to league average. Their batting average on balls in play even regressed to a league average number.

There are two caveats here, however:

1. There’s a bit of survivor bias; due to injury, Jake Arrieta didn’t pitch in the second half in 2012, for example. Joe Blanton was on this list twice but was released once by the Angels after his poor early season in 2013. Pitchers who might not have regressed, for one reason or another, could have been prohibited from giving that non-regressing performance, thus skewing the numbers a bit.

2. The group’s results improved significantly, but they still underperformed their second xFIP by three-tenths of a run, and their second half xFIP was worse than their first half xFIP. In other words, players that underperform to this degree for half a season aren’t likely to continue to underperform to that same level, but they might be likely to still underperform, and you shouldn’t expect their second half performance to match their first half peripherals.

Of course, the Yankees would probably take a 4.30 ERA from their new pitcher, who might also welcome that number after his rough start. Especially given that the cost was minimal, and McCarthy has a track record of success that suggests that a poor three months probably doesn’t mean that he’s now permanently terrible.

For one, McCarthy’s xFIP is the best on that list above. His combination of stellar walk and ground-ball rates is really only equalled by Rick Porcello in 2013. And it’s built on a compelling back story — McCarthy bulked up this offseason in an effort to have more staying power, and in return, his velocity increased to a career high.

And though McCarthy has been looking for a better change up ever since he switched to featuring the sinker, his home run problem this year hasn’t been a function of a platoon split. He’s had a better strikeout, walk and home-run rate against lefties than righties this year. Not that the split is sustainable — he still has slightly worse numbers against lefties than righties for his career — just that lefties can’t be blamed for his homer rate this year. Perhaps the 11 home runs he’s given up in the hitter friendly parks in Arizona and Colorado (versus the four he’s given up elsewhere) have a little more to do with the ledger standing as it does.

At least the Yankees and their home park — third-friendliest in the league to lefty power hitters — can hope so. They’ve got the rest of the (non-ERA) numbers on their side, it looks like.




Print This Post



Graphs: Baseball, Roto, Beer, brats (OK, no graphs for that...yet), repeat. Follow him on Twitter @enosarris.

40 Responses to “The Yankees Bet on Brandon McCarthy and xFIP”

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

    Amanda McCarthy in a pinstriped ….

    What did your article say again?

    Vote -1 Vote +1

  2. Kyle says:

    Brandon McCarthy, for his part, seems to be well aware of the concepts of FIP and xFIP. This brings up a topic I’ve always wondered this about with respect to DIPS theory. We all know that it’s obviously useful for a GM in predicting player performance. However, what happens when a player starts to think to himself, “I’m going to disregard my own BABIP and HR/FB numbers, because they don’t influence my xFIP”? Could that change in perspective actually influence the gap between peripheral predictions and runs allowed? I have no idea whether that is true or not. If anyone knows if there has been research done on this please direct me to it.

    Vote -1 Vote +1

    • Catoblepas says:

      I know Zack Grienke talked about his FIP specifically — THIS article from earlier this year mentions it, and links to the original quote. Basically, it seems like focusing on it wasn’t very helpful for him, as controlling quality of contact (ground balls, dingers) wasn’t something he was able to do. I don’t know of anyone focusing just on xFIP and ignoring HR/FB, though.

      Vote -1 Vote +1

    • DanUggla'sRetirment says:

      What exactly are you asking for?

      Vote -1 Vote +1

      • Kyle says:

        He’s a guy who’s aware of the research and also seems to be lagging in ERA-xFIP… I’m asking because I wonder if those are connected. Was that not clear?

        DIPS shows strong correlations and predictions, but I don’t think that it claims any causal link between peripherals and ERA. A pitcher thinking of peripheral stats as a thing to optimize might not help improve his ERA. Of course, I don’t know how McCarthy looks at his stats when he determines his process.

        Vote -1 Vote +1

        • The Legend of Bagger Vance Law says:

          I think your own question holds the answer.

          Vote -1 Vote +1

        • mattpullman says:

          Just think about if you were a pitcher… you are essentially striving to limit your baserunners (and runs, of course, but runs can’t score if nobody ever reaches base).

          In each individual plate appearance, a pitcher is attempting to get that player out. Walks are a big no-no. You’re taking “luck” out of the equation if you walk someone. In order to avoid walking someone, you need to throw strikes (unless of course your competition is dreadful and will consistently swing at balls).

          So if you need to throw strikes, you’re attempting to throw strikes that will either induce weak contact or no contact at all (whiffs and taken strikes). Inducing weak contact inherently lowers your BABIP and HR/FB rate. Inducing no contact takes luck completely out of the equation, but in your favor.

          Trying to ‘pitch to FIP’ as opposed to simply winning the game at hand is a pretty silly notion. A successful pitcher will have a low FIP as a result of them trying minimize baserunners and win the current ballgame.

          Vote -1 Vote +1

        • Kyle says:

          @mattpullman:

          I see what you’re saying, and I agree that there’s value in the peripheral stats, it’s just not everything, and a lot of it is based on assumption. Take xFIP. xFIP essentially says that a pitcher does not control his HR/FB rate. This is observably true from a statistical POV, because players in the current environment tend regress to a certain HR/FB rate. However, it’s true a-posteriori, not a-priori.

          It’s obviously true that pitchers do not want to allow HRs. However, a pitcher could read about xFIP and say, “I’m going to adjust my strategy to try and deny /all/ fly balls, and I’m not going to sweat the homers.” Might wind up looking better on paper, but does that really make him a better pitcher?

          I’m just speculating for the most part. I don’t know what a no-flyball strategy would even look like. I happen to think that with DIPS there is an assumption about the game theory and strategy of the game being stable. I think if more players are aware of the way they are valued, it will change the dynamics of the game.

          Vote -1 Vote +1

      • RC says:

        “DIPS shows strong correlations and predictions”

        The correlation for FIP is barely stronger than the one for ERA, and only in pitcher-friendly parks (in hitter friendly parks, its less predictive)

        Vote -1 Vote +1

    • Stan Gable says:

      I’d say it would be of help if a pitcher was more similar to Zack Greinke than Brandon McCarthy as far as talent goes.

      Vote -1 Vote +1

    • rustydude says:

      Interesting question. It sort of mirrors the physics theory that the act of observing a phenomenon will invariably affect the phenomenon – or something like that. In this case, a pitcher knowing DIPS theory is bound to have an effect, one that might not be possible to identify or measure, but still might exist.

      Vote -1 Vote +1

      • heisenberg says:

        I think u meant my theory, but it’s more like ‘by the time you observe and conceive some minuscule thing, its properties have already changed’

        Vote -1 Vote +1

  3. cnote66 says:

    Does Nuno move into the D-backs rotation or the pen?

    Vote -1 Vote +1

  4. Chris says:

    Watching games, it seems McCarthy gets really hit hard from the 4th inning on. He can cruise masterfully the first time through the order, but then he can get tied off on in the 4th and 5th innings. I am not sure how much this can improved.

    Looking at the numbers from Baseball Reference to see if this is a trend:
    2014 McCarthy-
    Innings 1 to 3: ERA=3.00, BAA=.241
    Innings 4 to 6: ERA=6.51, BAA=.351

    We would expect an increase, but this is quite a jump. Looking at somebody else performing below their peripherals, I grabbed Homer Bailey as a comparison.
    2014 Bailey-
    Innings 1 to 3: ERA=3.83, BAA=.241
    Innings 4 to 6: ERA=4.60, BAA=.299

    Very interesting similarities between these two for the first 3 innings. However over the next 3 innings we see a much more severe drop for McCarthy vs Bailey.

    More analysis is needed, but it seems like McCarthy has the stuff to mow threw the order once. But it starts appearing any deception quickly disappears the second time around, and he is getting knocked around.

    McCarthy could be a very good pitcher, but might be more effective in a relief role than as a starter.

    Vote -1 Vote +1

    • Ryan L. says:

      And the reason for his struggle from the third inning on is simple. He just lacks deception. He throws three different fastballs in his four seamer, sinker, and cutter, ranging from 89 to 94 MPH. Other than that, he relies on a 78 to 82 MPH curve ball. His ever tinkering attempt at throwing a change up has increased the velocity of his change from 82 last year to 87 this year, likely meaning that he isn’t even throwing a change anymore. Those mislabeled changes are probably just cutters that never cut.

      There are very few starters I’m the league who can get away only having a fastball/slider or fastball/curve. Ervin Santana did for a while until it really caught up with him in 2012. He changed his change up grip starting last year, which saw him significantly improve over his 2012 numbers. This year, he is throwing his change twice as mmuch as usual and is having one of his better seasons.

      Point is that McCarthy will never live up to his stellar peripherals without developing a change-up or splitter that is at least average. If he can’t then perhaps he could be a very valuable arm out of the bullpen which could then possible allow the Yankees to move Warren back into the rotation.

      Warren is dealing with the exact opposite problem. He has had trouble simplifying his repertoire for bullpen work. Most relievers are almost exclusively two pitch guys….Just look at Betances and Robertson. Warren, on the other hand, has five pitches he throws regularly. He has a four seamer, two seamer, a curve that is anywhere from average to slightly above, and two plus pitches in his slider and change….The slider may even be plus plus.

      Warren was brought up as a starter. However, he has been so spectacular out of the pen that the Yankees have been cautious about moving him back into the rotation….Especially because the starting rotation has been so awful this year that the Yankees need all of the quality bullpen arms that they can get.

      Vote -1 Vote +1

      • m_pemulis says:

        “the Yankees have been cautious about moving him back into the rotation….Especially because the starting rotation has been so awful this year that the Yankees need all of the quality bullpen arms that they can get.”


        Vote -1 Vote +1

  5. Emcee Peepants says:

    I commented on this earlier in the year, and it seems to have held true as the season has gone on, but McCarthy appears to either really have “it” in a given start or really does not, to a pretty extreme degree. He has 8 “bad” starts giving up 5 or more ER, 7 “excellent” giving up 2 ER or fewer, and only 3 “good” starts with 3 or 4 ER (actually none with 4). His lack of consistency or predictable performance is not something one would seem to want on a potential contender (which makes me very happy he went to the Yankees).

    Vote -1 Vote +1

  6. Dino says:

    Am I missing something about Nuno or does anyone else think that AZ could have gotten a bit more in this deal?

    Vote -1 Vote +1

  7. Bryan says:

    Brandon McCarthy is just a better pitcher than Vidal Nuno; no argument there. And his xFIP-ERA is all out of whack. But does Cashman (or anyone else) think that putting McCarthy and his great GB% in front of the worst infield in baseball is going to improve his performances?

    Vote -1 Vote +1

    • Cliff says:

      The diamondbacks are actually worse defensively than the Yankees

      Vote -1 Vote +1

      • Bryan says:

        Looking at the combined UZR of their current starting infields I see that the Diamondbacks at -2.7 and the Yankees at -5.5. Am I missing something?

        Vote -1 Vote +1

      • Phantom Stranger says:

        Yep, even the gold glovers on the team aren’t having great years in the field. Bad defense in a hitter’s park like Arizona spells trouble for pitchers that rely on groundballs.

        Vote -1 Vote +1

  8. The Legend of Bagger Vance Law says:

    McCarthy’s xFIP will raise as his strikeout rate decreases with the move to the AL and his FB and BB rate should increase. No matter what he should be an instant upgrade over Nuno for the time being, so you can’t blame the Yankees for making the move. I would worry that the Yankees entire staff is now too right handed. They are only carrying two lefties and Toronto and Baltimore rake against RHP. I don’t think he is inconsistent, for the most part his excellent starts have been against the poorer hitting NL teams and his bad starts against the better hitting teams. I would even give Clay Buchholz a shot at less than 3 earned and 7 innings against the Padres.

    Vote -1 Vote +1

    • Wobatus says:

      Orioles are at 102 wRC+ against righties, above average but not great. They are actually a little better against lefties at 105.

      I was a little surprised that the Jays are actually so much better against righties than lefties, 115 to 90 wRC+. Their big bats, Bautista, EE and Cabrera are all righties. But those guys can all hit righties too, but Lind, Francico, Rasmus, Reyes have all been worse to much worse against lefties.

      Also, the yankees xFIP against lefties is 3.53, 3rd best in baseball. Against righties they are at 3.66, 16th. Betances, Tanaka and Robertson don’t appear to have platoon issues, at least this year, and Kelley and Warren seem fine in that regard as well this year.

      Vote -1 Vote +1

  9. Hmmm says:

    More than anyone else I have seen, when Brandon McCarthy misses with a pitch, it gets drilled for a home run. From what I’ve seen, his misses tend to be absolute thigh-high meatballs which seem to turn even the weakest batter into a slugger.

    Vote -1 Vote +1

  10. I don't care what anyone says:

    Did Arizona confuse Vidal Nuno with Aldo Nova?

    +5 Vote -1 Vote +1

  11. Izzy says:

    Why would anyone trade for Vidal Nuno? He seems like the kind of guy that almost any team could replace with someone from AAA and not miss a beat. I can’t imagine why you would want to give up an asset for him.

    Vote -1 Vote +1

  12. Grant says:

    If only the Yankees realized how bad their infield defense is…

    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>

Current day month ye@r *