IMPORTANT BLOG ANNOUNCEMENT

PLEASE CHANGE YOUR BOOKMARKS AND FEEDS TO THE NEW URL, YANKEEANALYSTS.COM. TYU IS IN NO WAY AFFILIATED WITH THE NEW YORK YANKEES OR YANKEES UNIVERSE.
Jun 152010

Andy Pettitte is having a good year.  In 12 starts and 80 innings he has won 8 games and lost only 1, given up 24 earned runs, struck out 55 and walked 22.  He has an ERA of 2.46, a K/9 of 6.16 and a BB/9 of 2.46.  Whether it is because of the low ERA or the visually appealing win-loss record, some Yankee fans have deemed Pettitte the Yankees’ best starter and have called for him to be considered for the All-Star game this year.  With all due respect to Pettitte, and to his many fans (myself included), this would be a mistake.

Let’s get the stats out of the way. Pettitte’s ERA is artificially low.  His BABIP is a career-low .256 and his FIP is 3.73.  His batted ball data is right in line with past results, so it’s not as if he is suddenly generating weaker contact.  One thing that is slightly out of line with past results is his HR/FB ratio is 8%, slightly off his career average of 9.8%.  His strand rate of 82% is even more out of whack with his career average of 71%.  As a result, Pettitte’s xFIP stands at 4.08. What does this mean? It means that the statistics that give us the better indications of how well a pitcher did at controlling the things he can control tell us that Pettitte has pitched like a 3.75-4.00 ERA pitcher.  Are xFIP and FIP the end-all determinant of the quality of a pitcher?  Is having a top-8 FIP a requirement for inclusion into the All-Star Game?  Obviously not.  But FIP and xFIP can give us a better indication of how well Pettitte has truly done so far this year, provided you keep in mind the vagaries of small sample sizes.

Contextualizing Pettitte’s year in terms of advanced data and comparing him to other AL starters leads me to conclude that while he is having a great year so far, he doesn’t belong in the All-Star Game. In the American League in 2010, 13 pitchers have a better FIP than Andy Pettitte.  Leading the way is Cliff Lee with a 1.92 (albeit in a smaller sample than others), and behind him is Francisco Liriano with a 2.09 and one young Phil Hughes with 2.89.  If you prefer xFIP, 17 pitchers have a superior xFIP to Pettitte in 2010, including Phil Hughes with 3.69 and CC Sabathia with 3.92.  Leading the way is Francisco Liriano (2.94), followed by Cliff Lee (3.01) and Ricky Romero (3.21).

So where does this leave us with Pettitte?  Mike Axisa of RiverAveBlues looked at his start to the season and concluded the following:

“Pettitte ripped a page out of the 1997 playbook this year and has been the Yanks’ best and most consistent starter…You don’t expect him to keep performing this well, but the season is close to 40% complete, and he just keeps doing it. Season totals of a .256 BABIP and 82.1% LOB% tell us to expect a regression, but I get the sense that we might be waiting a while. Sometimes unexplainable things happen to extraordinary players, which Pettitte certainly is.”

As Mike notes, regression is likely with Pettitte, but he is still having an incredible year.  Pitching to the 14th best FIP in the American League is no small task.  The fact that Pettitte is doing it at his age, after the career he has had, makes it all the more impressive. Expecting some regression as the year progresses should not diminish a full appreciation of his 2010 season and his career in general.  But Pettitte is not an ace for the Yankees this year, and he clearly he isn’t the Yankees best starter this year either.  The Yankees’ best starter has been Phil Hughes, and if anyone should represent the team at the All-Star Game this year, it should be him.

For the record, I would pick the following pitchers to fill out the 13 man staff:

SPs: Francisco Liriano (starter), Cliff Lee, Ricky Romero, James Shields, Jon Lester, Felix Hernandez, Jered Weaver and Phil Hughes.

RPs: Neftali Feliz, Joakim Soria, Mariano Rivera, Rafael Soriano and Jonathan Papelbon

45 Responses to “Andy Pettitte Is No All-Star”

  1. What if he’s 11-1 or 12-1 at ASB? All star worthy then?  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    If the other stats were roughly the same? No. I don’t use wins and losses as a determinant of pitcher skill. They hinge too heavily on run support and bullpen strength. I’d take Greinke over Pettitte for the ASG any day, and he’s 2-8.  (Quote)

    [Reply To This Comment]

    Adam Reply:

    Tyler Clippard of the Washington Nationals is 8-3. Sounds good right? Yeah, he’s a reliever. I’m with SR, wins don’t mean anything. It’s a pity that so often the commentary we hear on the radio, read in columns, or hear on TV is based around wins/losses and ERA. Nice article SR.  (Quote)

    [Reply To This Comment]

    Clintfsu813 Reply:

    You cant compare Clippard with Andy even though their W/L is similar. Andy has been loads better and as for Grienke..cant lie. I’m a little AL East biased. I never compare guys who pitch in another division with an AL East pitcher.  (Quote)

    [Reply To This Comment]

  2. I do think Hughes has been better overall than Pettitte, but unless he’s not going to pitch, I don’t want someone on an innings limit going to the ASG. I think you can make a case for Pettitte. I hope I’m not just being a Pettitte homer here, but both BB-Ref WAR and VORP have him as the second-most valuable pitcher in the AL. Those stats aren’t perfect, but neither are ERA or FIP.

    I think he’s going, though, especially with Girardi picking the pitchers. Can’t say I mind it. If Wakefield could go last year with absolutely horrible peripherals…  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    w/r/t hughes, i definitely wouldn’t want him in the ASG as a fan. i wouldn’t want to waste the pitches/innings, and i’d be terrified that he’d get hurt.  (Quote)

    [Reply To This Comment]

    bexy Reply:

    Yep. I’m not really terrified of them getting hurt, they could get hurt in any start, they could get hurt walking down the street, but don’t waste the innings.

    I think Weaver should start. I know IT COUNTS!!!! but the ASG is enough of an exhibition game that you can mix objectivity and subjectivity. He’s quietly putting together a very good season and he’s the hometown guy.  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    Yeah. Rationally, he could get injured in his first start back from the ASG, or any other start. Irrationally, I would just drive myself crazy with anger if he hurt his shoulder warming up to come in in the 7th inning of the ASG.

    Didn’t even think about the Weaver/hometown connection. I think he should start too.  (Quote)

    [Reply To This Comment]

  3. Wait, I agree on your take using wins as a determining fact for the the All-star game, but when did ERA and WHIP become meaningless or not All-star worthy determining statistics? I understand the premise behind the blog post, but if you’re final conclusion states that James Shields (4.55 ERA and 1.41 WHIP) is a more deserving player then Andy Pettitte (2.46 ERA and 1.10 WHIP) then I have to question the validity of the entire post…it smells like a fabricated controversial post to generate a response.  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    I thought including Shields might generate a question.

    First things first: I didn’t mention WHIP in the post at all. I don’t have a problem with it.

    While I do think that ERA is a far better stat than W-L, it does have a few weaknesses. When you use FIP and xFIP you see things that you wouldn’t normally see by simply looking at ERA. Of course, you have to look at batted ball data, strand rates, hr rates, etc, to get a full picture, but the whole idea behind using stats other than W-L/ERA is to try to paint a fuller picture of a pitcher’s performance.

    That said, Shields has had a very good year. His FIP is 3.88 and his xFIP is 3.38. His K/9 rate is elite, as is his K/BB ratio. As you can see in the link below, he’s had two horrific outings and some very very good outings. He’s not the AL’s best pitcher, but his ERA might lead you to underrate him a bit.

    http://www.baseball-reference.com/players/gl.cgi?n1=shielja02&t=p&year=2010  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    You know, I think maybe this might deserve its own post tomorrow. Maybe I could go over the top 20 AL pitchers in 2010 and rank them by my standards, and then we could have a discussion about it.  (Quote)

    [Reply To This Comment]

    Moshe Mandel Reply:

    That would be great.  (Quote)

    [Reply To This Comment]

    eddieperez23 Reply:

    Yes…that would be an interesting read…more fodder for debate! :)   (Quote)

    [Reply To This Comment]

  4. Now this is simply over-utilizing SABER-metrics. All Star participants are determined (for whatever reason) based on results from 25-45% of the season. Whether a player has been “lucky” may affect their stats and over the course of a full season they may not stay in line with their numbers, but for better or worse that is not what the all-star game is about. It’s about seeing the players who got the results, and results are reflected in the main stream stats such as W-L and ERA. In evaluating whether someones year was really as good as their line you can make that arugment but for a popularity and small sample size selection like the All-star game you have to go solely off of results and not projections or luck.
    That said whether Andy (or Phil) is or is not an all-star can be debated but I dont think trying to determine whether he will regress in the 2nd half is at all relevant to the discussion.  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    “It’s about seeing the players who got the results, and results are reflected in the main stream stats such as W-L and ERA. In evaluating whether someones year was really as good as their line you can make that arugment but for a popularity and small sample size selection like the All-star game you have to go solely off of results and not projections or luck.”

    Just because stats like W-L and ERA are popular doesn’t mean that they’re qualitatively better than others. While I would agree that you have to account for small sample sizes when evaluating FIP, xFIP and BABIP, I would argue that you want the “best” players in the ASG, and “mainstream” stats like W-L and ERA don’t necessarily reflect who has been the best.  (Quote)

    [Reply To This Comment]

    lenNY's Yankees Reply:

    Valid points in your article. My only problem with it is my trust in FIP and xFIP. Ubaldo Jimenez, who is having a Bob Gibson-type year, loses rank big-time in FIP and xFIP.

    http://www.fangraphs.com/leaders.aspx?pos=all&stats=pit&lg=all&qual=y&type=1&season=2010&month=0

    Frankly, I’m not sure if I’d rather have anyone else on the mound this year than Ubaldo.

    As far as I’m concerned, Pettitte, Hughes and Rivera are all All-Stars.  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    I definitely see what you’re saying about Jimenez. Is it possible, though, that he’s gotten just ever-so-slightly lucky this year? His strand rate is 91% and his HR/FB% is 4%.  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    Of course, what I said shouldn’t take away from Ubaldo’s greatness this year. Anyone with an ERA of 1.1 is going to have some kind of FIP discrepancy. More than anything, I think it’s good that we use a broad range of statistics, and not rely on any one metric to tell us “who’s good”.  (Quote)

    [Reply To This Comment]

    Moshe Mandel Reply:

    This is argument that I hear a lot, and I’m not sure where I stand. The advanced metrics try and take luck and defense out of the equation, to get at the core of a pitcher’s performance. In regard to predictive value, they certainly win out. However, I’m not so sure in terms of evaluating who was better in the past. Meaning, if someone is more dominant, and needs his fielders less, does that make him objectively better? It makes him more talented and more likely to suceed in the future, but does it mean he has been better to this point? I grapple with this a lot, and the answer is not a simple one. There is some value in actual results and performance rather than expected results and performance when looking retrospectively. I happen to lean towards your side of the fence, SR, but I think the other side of the argument has some validity to it.  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    Two things:

    1. If someone is more dominant, needs his fielders less, walks very few guys and strikeouts a lot, and has a 3.50 FIP and a 4.25 ERA, then I would argue that yes, he’s objectively better than they guy with the 3.50 ERA and the 4.25 FIP. That’s a generality fraught with peril, but I think the core of what makes a pitcher good or dominant is his ability to keep guys off base (BB), keep guys from putting the ball in play (Ks), and to a lesser extent keep the ball on the ground.

    2. More to a general philosophical point – I think there’s a danger in putting a false dichotomy between “actual results and performance” and “expected results and performance” when talking about ERA v. FIP and xFIP. FIP and xFIP aren’t just projections, they’re stats too, stats that measure “actual results”. ERA carries this weight to it, because we’ve had it for so long, because people 40 years ago used it, because it’s easy to understand and part of our national conversation, but it’s no more “actual results” than FIP is. FIP is not perfect, by any means, but I think it’s a bit misleading to call one stat “real” and the other “projection”. If we redefine what we consider “good” and “real” over the next 20 years, people may look at FIP (or FIP’s replacement, whatever that may be) as the “actual result”.  (Quote)

    [Reply To This Comment]

    Moshe Mandel Reply:

    Here is the problem- the stat tries to distill results entirely attributable to the pitcher. By doing so, it removes a range of outcomes that were part of the real and actual historical event. You are right, I created a dichotomy that was unclear and unfair, at least with regard to FIP (I think you would agree that xFIP is about expected results and performance). More accurately, I think I could fathom a reasonable argument for including those fielding dependent events as part of the record when considering rewarding past performance. I would not agree with that argument in most cases, but I would certainly consider it a valid opinion and one with some merit.  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    Yes to all that. I’m not sure where I come down on it, but it’s certainly reasonable. Hopefully I didn’t create the impression that I regard ERA as equivalent to W-L, because I don’t.  (Quote)

    [Reply To This Comment]

    Joe O Reply:

    Glad I stirred up some debate and have Moshe backing me up (and then going the other way — we can discuss this on the train sometime if either of us manage to stay awake) somewhat. What it comes down to for me is that for an all-star game we reward popularity and easily quantifiable results for the average fan. For a HOF debate I would tend to side with you as well, but given the nature of the game, the sample size, the average fan voting — when you make an argument of deserving to be in an all-star game I think ERA gets far more weight than FIP and other “new stats”.  (Quote)

  5. Just tryin to stir it up? You cannot be serious. Cliff Lee?  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    What about Cliff Lee and his 15.00 K/BB ratio do you find to be absurd?  (Quote)

    [Reply To This Comment]

    the other Steve S. Reply:

    What about Andy’s 8-1 record do you find absurd?  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    Boy, if we have to review why Wins and Losses are a bad way to evaluate how good an individual pitcher is, then we will just have to start from the beginning.  (Quote)

    [Reply To This Comment]

    the other Steve S. Reply:

    Well, if you think ‘doing well’ is somehow more important than winning games. I think we’re done here. I expect Andy will be an All-Star and will acquit himself well. You pick the stats you think matter and I’ll pick mine. It’s all opinion and mine is just as valid as yours. I don’t remember hearing any discussions about anybody getting into the HOF because he had a great FIP, but 300 wins is sure a ticket.  (Quote)

    [Reply To This Comment]

    Moshe Mandel Reply:

    That doesn’t make it right. Listen, the win isn’t irrelevant, but it isnt very helpful either. Greinke is a perfect example of that, as he has pitched pretty well yet has, what, 2 wins? Meanwhile, Tyler Clippard has vultured 8 wins somehow.  (Quote)

    [Reply To This Comment]

  6. 2.46 ERA. That’s All-Star material. All the peripheral stats are interesting, but the ability to bear down when the pressure is on means something, like ERA and WINS. Winning baseball games, not “stuff”, or even great stats is what MLB pitchers try to do in the real world, not in Fantasy Land….  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    Couldn’t disagree more with the entirety of this comment. As demonstrated in this post, Pettitte’s 2.46 ERA is deceptive. As demonstrated here, there, and everywhere, Wins are a dumb stat and don’t tell us much of anything about a pitcher’s skill. Finally, “the ability to bear down when the pressure is on” may be a real thing, but it’s certainly not evidenced in ERA and Wins, as you seem to suggest.  (Quote)

    [Reply To This Comment]

  7. Why Papelbon? What was your process for selecting relievers? I’m guessing that it didn’t involve FIP or xFIP or WAR.  (Quote)

    [Reply To This Comment]

    kmarx Reply:

    Jonathan Papelbon:

    ERA: 2.88
    FIP: 5.07
    xFIP: 4.90

    WAR: -0.2  (Quote)

    [Reply To This Comment]

    bexy Reply:

    Very good point. I’d be interested to hear this explanation.  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    Yeah, honestly I didn’t spend nearly enough time looking at relievers. It should have been Rauch over Papelbon, no doubt.  (Quote)

    [Reply To This Comment]

  8. Well framed Joe O and Moshe in distinguishing between actual performance to date and predicted performance based on advanced metrics.The All-Star game, at least for the reserves and pitchers, is supposed to reward those who have had the best results during the first half of each season, regardless of how the results may have been shaped by anomalous of BABIP,xFIP, strand rates,HR/ fly ball ratios etc. The fact that someone may have had “good fortune” in the first half as measured by results incongruent with these expectancies, should not be disregarded. or are we saying that someone like Jose Bautista should not be given serious consideration because a. he has been historically mediocre, b. his HR/flyball % is out of whack with historical precedent and therefore lucky c. we “know” that eventually he will regress to a much lower level? T o take it a step further, are we saying that the winner of the game should not be the team who scored the most runs but rather the team that actually hit the ball hardest, that a 400 foot drive caught at the wall should be given more credit than a bases loaded broken bat single that falls in? Finally, are we saying that the pennant winner should not be determined by the team that actually won the most games but rather by the team that “should” have won the most games as determined by a multi-level expected runs scored and allowed/pythagorean analysis?
    Of course not. By its very nature, baseball results, for both players and teams, will never perfectly represent who actually played “better” in particular game or for a particular period.

    Finally, in picking an all-star pitching staff does the name David Pice ring a bell? James Shields over David Price? Please. A perfect example of sabermetrics divorced from reality on the field.  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    “Finally, are we saying that the pennant winner should not be determined by the team that actually won the most games but rather by the team that “should” have won the most games as determined by a multi-level expected runs scored and allowed/pythagorean analysis?”

    Yes, that’s exactly what I’m saying. Obviously. I want the games to be played in spreadsheets, not on the field, with calculators and fancy formulas and nerds with glasses.

    /annoyed sarcasm  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    And yes, the name David Price rings a bell. If you were able to look past any stat other than his ERA and his W-L record, you might realize that he’s not in the top 10 of AL pitchers. If that assaults some sort of sensibility about “what’s real” to you then I don’t really know what to say.  (Quote)

    [Reply To This Comment]

    classicsteve Reply:

    “And yes, the name David Price rings a bell. If you were able to look past any stat other than his ERA and his W-L record, you might realize that he’s not in the top 10 of AL pitchers”

    But he has been better than James Shields, a pitcher on the same staff who you believe should be an all- star, and presumably, with minor variations, the beneficiary of similar defensive support.A quick, admittedly incomplete statistical comparison:

    W-L ERA H/9 HR/9 WHIP K/9 W/9 BABIP

    Shields 5-5 4.55 10.7 1.4 1.412 8.7 2.0 .344

    Price 10-1 2.31 7.4 .6 1.226 6.7 3.7 .260

    Contrary to what you suggest, Price has far better numbers when looking at H/9, WHP and HR/9. The disparity in BABIP is high and should converge as the season continues.Agreed. However, Shield’s superiority in K/9, W/9, and by extrapolation K/W does not necessarily mean that he has pitched better. While his CONTROL as measured by the walks, has been better, it does NOT necessarily mean that his COMMAND within the strike zone has been better than Price’s. While he has been more succcessful putting batters away by the K,It is entirely possible that he has been making more mistakes- putting more pitches in better hitting zones- and this has been, at least in part the reason for the difference in BABIP between the two.  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    This is a decent response. In about 10 minutes I have a post going up about the top 15 pitchers, and in it I’ve moved Shields down from 8 to 13. I have Price at 15, but I could buy the argument for swapping them.  (Quote)

    [Reply To This Comment]

  9. I love saber staff, but (just like the Cy Young and the MVP) I’m not in favor of suddenly changing the standards currently being used. Guys are picked for these things based on things like RBI, saves, wins, and BA. These days things like era, WHIP, and OPS are used a lot more than even a decade ago, but using things like UZR, BABIP, and FIP for an all-star game is taking it a bit too far.
    (Especially with some of these stats that I still don’t put a lot of stock in, like UZR and Line-drive rate. Who decides which balls are line drives, anyway? Still sounds a lot like judging figure skating to me.)  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    I think they probably decide what a line drive is the same way that a scorer decides what a hit and what an error is. Which, you know, is the way we get our idea of “earned runs”.  (Quote)

    [Reply To This Comment]

    oldpep Reply:

    In other words, it’s often decided incorrectly.

    Why the sarcasm? Is it because ‘line drive rate’ is obviously subjective, but used by you (and far too many saber guys) as if it were objective? I call BS on using any subjective stat as if it were objective. Line drive rate, is based on some observers opinion, and subject to all of his or her prejudices. Robbie Cano vs UZR shows just how little objectivity there is with these type of things.  (Quote)

    [Reply To This Comment]

    Stephen R. Reply:

    Your problem with line drive rate is that it’s subjective, correct? Do you have a problem with relying on the concept of unearned and earned runs as well?  (Quote)

    [Reply To This Comment]

Leave a Reply

(required)

(required)

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>

© 2011 TYU Suffusion theme by Sayontan Sinha