What the Good Book says: Yankee pitchers

On Jeter and his improved defense
Meet the new Boss, not nearly the same as the old Boss

With the arrival of Baseball Prospectus 2010 at my doorstep on Friday, I spent some time yesterday look at the Yankee hitters and how PECOTA projects them to perform this season. While these are rough estimates of the team’s performance, they’re indicative of the trends many experts expect to see emerge from the Bronx.

After highlighting the Yanks’ offense yesterday, I’m going to look at the pitchers today. For the starters, I’ll present W-L, IP, K, BB, ERA and SIERA projections. For more on SIERA, read through the five-part explanation at BP or Joe’s introduction from the 17th. Let’s dive in.

CC Sabathia — 15-10, 219 IP, 183 K, 55 BB, 3.66 ERA, 3.58 SIERA
As expected, the Baseball Prospectus experts highlight Sabathia’s workload. He’s a horse who “works deep into games.” He has “handled it before and seems a safe bet to do again.” He’s projected to start 32 games this year, and his win-total projections seem to be a handful on the low side. I’d expect nothing but a good season from CC this year.

A.J. Burnett — 11-11, 193 IP, 175 K, 81 BB, 4.57 ERA, 3.85 SIERA
PECOTA is always bearish on pitchers, and Burnett’s line here is no exception. If the Yanks’ second starter is pitching to a .500 record and a 4.57 ERA, the team might be in trouble. BP calls “the biggest triumph of his season” the fact that he stayed healthy and focuses on the disparities between what we call Good A.J. and Bad A.J. Depending upon how one characterizes it, Good A.J.’s ERA is more than five runs per game lower than Bad A.J.’s. He’ll probably soon be greatly overpaid.

Andy Pettitte — 10-11, 180 IP, 116 K, 64 BB, 4.70 ERA, 4.58 SIERA
BP’s final sentence on Andy Pettitte sums up my expectations for Andy’s 2010 campaign. “Pettitte seems a safe bet to give the Yankees another year of solid if unspectacular keep-’em-in-the-game pitching.” He’ll turn 38 this year, and his velocity has begun to dip as any 38 year old pitcher’s does. I always wonder if this is the year that age and a cranky elbow keep Pettitte down, but the Yanks have the depth to weather that happening.

Javier Vazquez — 14-11, 203 IP, 180 K, 54 BB, 3.85 ERA, 3.47 SIERA
I have him listed as the Yanks’ fourth starter, but if he duplicates his PECOTA projection, he would be in line to serve as the Yanks’ second starter this season. BP predicts an uptick in home runs allowed, but he’s a good bet for a lot of innings and a good number of strike outs. As BP says, at the price it cost the Yanks to land him, his return to the AL was a risk “very much worth taking.”

Joba Chamberlain — 9-10, 159 IP, 150 K, 67 BB, 4.45 ERA, 3.75 SIERA
When it comes to Joba Chamberlain, whoever wrote BP’s capsule reviews did not mince words. “What a mess,” they say. “It’s possible that no pitcher in the history of baseball has suffered through as many team-inflicted head games as Chamerlain has.” As I wrote on Friday, I find that to be a load of poppycock. It’s the media’s fault, although the Yanks shoulder some of the blame. BP expects Joba to wind up in bullpen with the arrival of Javier Vazquez, but someone has to hold down the fifth starter job. It’s Joba’s to lose.

Phil Hughes — 7-5, 103 IP, 90 K, 41 BB, 4.07 ERA, 3.90 SIERA
PECOTA has a problem with young pitchers who don’t have a long record of Major League pitching and have been wildly inconsistent. The system projects just 19 starts and 34 appearances for Hughes, but unless he gets injured or the Yanks pigeonhole him into the bullpen, he’ll far exceed 103 innings this year. The Yanks need Hughes to be more than a one-inning pitcher at this point in his career, and where he ends up out of Spring Training is anyone’s guess.

Mariano Rivera — 4-3, 57 IP, 54 K, 15 BB, 3.53 ERA, 3.19 SIERA
When it comes to Mariano Rivera and PECOTA, I just sit back and laugh. The Good Book has him pegged for 22 saves in 58 games. Rivera’s career low in saves is 28, and the BP Experts freely admit the problematic projection. In the past, they have termed Rivera to be “otherworldy,” and they note that he has “shut down everyone else, so why not Father Time?” At some point, he’ll decline or he’ll just retire. That point probably won’t be now.

Beyond those pitchers, it’s not really worth delving into the projections. BP likes Damaso Marte‘s chances for a rebound year and David Robertson‘s role as a potential set-up guy while PECOTA is cool on departed Yankees Phil Coke and Brian Bruney or Chien-Ming Wang‘s chances of recovery. The bullpen remains a bit volatile for now, but that’s the way the Yankees like it. They have enough young arms to plug the gap.

I personally have never been too in love with the way PECOTA projects pitchers across the board. The system seems to consistently underproject the top starters, and I’m not just saying that because the Yanks’ pitching numbers aren’t as stellar as we would like to be. There’s a certain volatility in projecting pitchers because they can be so injury-prone and much of what they do depends upon the defense behind him. Still, this is food for thought on a Sunday afternoon in February.

On Jeter and his improved defense
Meet the new Boss, not nearly the same as the old Boss
  • AndrewYF

    “What a mess”

    Ironic, coming from BP, owner and operator of the now highly-inaccurate PECOTA projection system.

  • kenthadley

    doesn’t look like a banner year for us if this comes to pass…..guess the Good Book doesn’t like the Yanks in 2010….think they’ll be taken by surprise by either Phil or Joba….young pitchers have a tendency to “break out”….see Halladay, Maddox, Glavine, etc….

    • http://www.secondavenuesagas.com Benjamin Kabak

      It’s funny though because as writers, they recognize that the Yanks have a great chance at repeating. The team write-up says that the Yankees are “in a very good position to repeat.” Whoever wrote it also appreciates Cashman’s team-building this year and his unwillingness to overpay Matt Holliday or Jason Bay. This iteration of PECOTA just disagrees.

  • http://www.facebook.com/pages/Bring-Melvin-To-America/193013541601?ref=sgm Andy In Sunny Daytona

    They have the starting five pegged to be a combined 6 games over .500? I’ll take the over.
    AISD Projects that the 2011 Baseball Prospectus to be worth $2.87

    • DCBX

      Chock full of win! :)

  • http://www.thesportsshowlive.com/ Joey H.

    Who said they are right anyway? I don’t agree with these numbers at all. Can we go back to last years projections and see how accurate they were?

    • http://www.secondavenuesagas.com Benjamin Kabak

      The projections are only right when it makes our team look good, right? :)

      At some point, I’ll haul out BP 2009 and check out the projections. One of the things about World Series winners though is that to get there, most players had to overperform their projections. See, for instance, the 2007 Rockies. A 2009 comparison would make for an interesting one-off post, but it doesn’t render judgment one way or another in PECOTA’s favor.

      • Mike F

        why dont you haul them out today? aren’t you curious?

        • http://www.secondavenuesagas.com Benjamin Kabak

          Because I have law school work that requires my attention today. If this were my full-time job, I’d be happy to get that post up, but until more of you donate, I’m stuck trying to make some money here.

          • Rob c

            I’ll see about getting some of my physician colleages to donate.
            The fewer lawyers practicng law the better!!

      • http://www.thesportsshowlive.com/ Joey H.

        I get what you mean. But you get what I mean as well. Only 15 wins out of CC and 11 out of Burnett is a bit chincy. I dont expect a 25 win projection for Sabathia. Maybe 17. Do these projections take run support into account? If so, they really think the offense fell off a cliff.

  • http://twitter.com/JamalG_BB Jamal G.

    All the essential metrics (i.e., non-win-loss-and-ERA numbers) seem to be within the realm of both their 2009 and career counterparts. They actually are projection the K:BB ratio of CC Sabathia (3.33) and A.J. Burnett (2.16) to both be above their 2009 marks (2.94; 2.01), and – in the case of the former – be above his career average (2.69).

    Honestly, with the way games play out with regards to the bullpen, are we really going to moan about win-loss projections for pitchers?

    • http://www.facebook.com/pages/Bring-Melvin-To-America/193013541601?ref=sgm Andy In Sunny Daytona

      I get what you are saying. They really shouldn’t even put the Win-Loss record on there.

      I’m still gonna moan about it though. So there. :)

      • http://twitter.com/JamalG_BB Jamal G.

        Hah, maybe “moan” was a tad too strong, but it has always befuddled me that people do gripe about that aspect of the projection so often.

        • bexarama

          I think it’s just because they’re a lot uglier than what we’re used to seeing from Yankee pitchers. Even the not-great SPs usually don’t have .500 or under .500 records because of the Yankee offense. But, of course, W-L is silly.

          • Lanny

            The same people that say w-l don’t matter are up in arms over a statistical projection of w-l’s. Comedy

            • bexarama

              Am I up in arms? No, Lanny.

  • Arman Tamzarian

    I’m curious as to what if anything can be done with AJ Burnett to provide some level of consistency? I freely admit I don’t know anything about pitching, but I imagine there’s got to be a way for him to recognize when he’s going bad, and make some adjustment. I don’t feel like I’ve seen many pitchers who are as night and day as he is on a regular basis

    • http://www.mystiqueandaura.com Steve H

      At 33 years old, unfortunately I think he is what he is. Cy Young somedays, Sidney Ponson on the others.

      • Arman Tamzarian

        I agree, I don’t totally understand PECOTA, but I imagine it must be near impossibly to project a player like Burnett. How do you project a player who’s variance is so wide on a consistent basis?

        • whozat

          at the end of the year, he’s not so different. K/9 of 8-9, BB/9 of roughly 4. ERA of roughly 4.

          • Arman Tamzarian

            Looks like your right. It seemed like he had a lot more 1-2 IP, 6 run games.

          • Reggie C.

            I’d hoped Burnett could lower his WHIP closer to his career average of 1.30, and he still could if he avoids running into the one-bad-inning a couple fewer times. Still, if AJ splits the difference to the tune of a 1.35 WHIP, the ranks of detractors would thin out.

            I sure hope Burnett remains an outlier to the Pecota system. Anything close to a 4.50 ERA would kill the pen down the stretch, and essentially make AJ the 5th man in the rotation.

      • Evilest Empire

        As long as he stays healthy and is good for 190+ IP, that’s a damn good pitcher to have. Even on his bad days, its generally just The One Bad Inning that kills him, and he’ll still give the team 6 innings on those days. At least you can put him against someone like Lester or Beckett (like, say, as his first start of the season) and like the Yankee’s chances.

  • Sweet Dick Willie

    While these are rough estimates of the team’s performance, they’re indicative of the trends many experts expect to see emerge from the Bronx.

    But…I heard, through a reliable source, that you just can’t predict baseball.

  • Jud

    I have a basic problem with BP projections and indeed with all similar types of scientific projections. The goal of these projections is to reduce the minimum total error of the projection. In order to do so, they introduce a healthy amount of regression to the mean. The unfortunate side effect of this system is that there is a compression of the results towards the mean. In other words if you’d look at the projected leaderboard of the top 50 players in WAR/WARP/VORP of these projections, you end up with a much more mediocre result as compared with the actual results of the real-life leaderboard. For example the CHONE projection has only 12 players with a WAR greater than 5, while in 2009 there were 24 such players.
    If it were up to me, I’d go about this backwards. Namely, I’d start with a constraint of a certain given leaderboard structure (e.g 25 players with WAR greater than 5). The projections job would be to assign these predetermined WAR values to specific players.
    Hope this wasn’t too long…

    • Accent Shallow

      This is an interesting point and should be addressed. It seems that the way projection systems are designed (with heavy regression to the mean) that they’re most likely to get someone like Raul Ibanez right . . . someone who’s an above average hitter, but not a superstar. (Of course, Ibanez is now what, 38? His aging curve is probably taking over) Young players breaking in and superstars are going to be the clear outliers.

      As to whether a more optimistic projection system would be more accurate, I have no idea. Everyone regresses to the mean, although it’s unfortunately not the same mean . . . A-Rod is a much different hitter than David Eckstein for example.

      • Jud

        I’m willing to concede the point that the more pessimistic projections are more accurate. My point is that I believe that these projections are not a “valid projection”. In order for a projection to be valid in my mind, it needs to be similar in it’s structure to the real life. By “similar in structure”, I’m referring to the distribution of the results.
        To take it to an extreme, both a projection that would have A-Rod with a WAR of 8 and a projection that would have Yuniesky Betancourt with a WAR of 8, are valid projections. One of them of course is reasonable, and the other extremely horrible, horrible – but valid.
        A projection system with NO ONE with a WAR of 8 is not valid as it does not model real life MLB results.

  • http://janeheller.mlblogs.com/yank_c.jpg T-Dubs

    I’m curious Ben, are the projections for Lee, Halladay, Lester, Felix, and Timmy in line with CC’s or is there a high outlier (i.e. a pitcher they project to have a far superior season to CC)?

    • http://www.secondavenuesagas.com Benjamin Kabak

      Lee: 12-11, 205 IP, 140 K, 56 BB, 4.12 ERA, 4.25 SIERA
      Halladay: 14-11, 218 IP, 160 K, 48 BB, 3.80 ERA, 3.78 SIERA
      Lester: 11-10, 190 IP, 152 K, 68 BB, 4.14 ERA, 4.02 SIERA
      Hernandez: 13-12, 221 IP, 190 K, 75 BB, 4.12 ERA, 3.60 SIERA
      Lincecum: 14-9, 204 IP, 219 K, 73 BB, 3.23 ERA, 3.10 SIERA

      Basically, it looks like PECOTA projects every top pitcher to have a worse 2010.

      • http://janeheller.mlblogs.com/yank_c.jpg T-Dubs

        Sort of what I figured. So basically CC would win the AL Cy Young with 15 wins and a 3.50 ERA. The voters’ heads would explode.

      • pat

        Hah, the best pitchers in all of baseball are projected to be 16 games over .500. Interesting.

        • Reggie C.

          There are simply way too many legit sucka-got-no-juice types in the majors who will almost always fail against the Greinke, Lincecum, Felix, and Lester level of ace. The elite pitchers will put up elite numbers.

      • bexarama

        holy crap. You weren’t kidding when you said they’re bearish on pitchers.

        • http://www.secondavenuesagas.com Benjamin Kabak

          Here’s another good one:

          Verlander 12-11, 208 IP, 179 K, 74 BB, ERA 3.99, SIERA 3.87

          Greinke’s projection is messed up because the book has the Royals’ pitchers innings well beyond anything reasonable. I’m not sure why the team’s pitchers are projected to throw 1000+ innings each. But the book projects 31 starts and 42 overall games for Greinke with 179 K, 55 BB, 3.71 ERA anda 3.47 SIERA.

          That’s just a weird weird projection.

          • http://www.facebook.com/pages/Bring-Melvin-To-America/193013541601?ref=sgm Andy In Sunny Daytona

            11 Pinch hit appearances. Have you seen that Royals offense?

          • bexarama

            But the book projects 31 starts and 42 overall games for Greinke

            GREINKE TO TEH 8TH!!11!!!!

      • http://www.mystiqueandaura.com Steve H

        If Lee and Felix combine to go 25-23, the Mariners will be lucky to win 65 games.

      • Accent Shallow

        Lee: 12-11, 205 IP, 140 K, 56 BB, 4.12 ERA, 4.25 SIERA

        Seems like a totally reasonable projection to me!

        /will keep saying it until Lee implodes

        • bexarama

          Heh. I like him, but I’m waiting for this too. People are acting like he’s Sandy Koufax since the 2009 postseason, and he’s really not.

      • Rey22

        I’d be willing to bet my everything there is absolutely no way Halladay gets that much worse moving out of the AL East into the NL. He’s no Barry Zito.

        …Barring injury of course.

  • Dan (Boston’s biggest NYY FAN)

    This really makes me sick.. I want to throw up.. seriously… how easy is it to just get every player and downgrade their stats and throw out a book and call it an expert analysis.. CBS SPortsline projections year to year are more accurate..

    So i guess The Yankees with their starters 6 games over .500 aren’t expected to win more then… 75-80 games this year???

    can someone kick Nate Silver in the face or whoever came up with this bullshit..

    im actually pretty upset now.. haha… really am. i need to go take a walk…

    • Jud

      Unfortunately, it seems that Nate is too busy with this side show of his (like accurately predicting presidential and congressional results…) and he is no longer involved in Pecota. I think that this transition to Clay Davenport is what is causing BP major problems.

    • Tom Zig

      Eh it’s just a projection. Projections are generally on the conservative side. It doesn’t bother me.

      • bexarama

        It bothered me for about two minutes, then I looked at their predictions for Mo and giggled, and then I saw their projections for other pitchers (see what Ben posted above for Greinke, Lee, Halladay, King Felix, etc.) and I felt much, much better.

    • Mike F

      i don’t think Nate is still involved in these projections, is he?

    • Lanny

      You really need to relax. It’s a projection designed to sell books and get people talking

  • http://www.facebook.com/pages/Bring-Melvin-To-America/193013541601?ref=sgm Andy In Sunny Daytona

    So Ben, BP isn’t necessarily projecting the man against himself, but the history of players that projected like him at the players current age?

    Does that even make sense?

    • bexarama

      I’m not Ben obviously but I’m pretty sure that’s exactly what they do.

      • http://www.facebook.com/pages/Bring-Melvin-To-America/193013541601?ref=sgm Andy In Sunny Daytona

        Doesn’t matter if you’re Ben or bexy (ht JMK), I’m glad that I have a better understanding. It has to be difficult to project HOF-type players, because the reason they are Hall bound is because they are so much better than the average.

        BTW, I’m deaf from the bell that went off when I finally understood.

        • bexarama

          Yeah. If they could project Mo, well then, he wouldn’t be Mo, you know?

  • Dan (Boston’s biggest NYY FAN)

    Derek Jeter a .286 hitter… ohhh boy we are in trouble!! hahaha

    fuck nate silver and his rip off book. espn free draft kid is good enough for me…

    quest for 28

    • http://www.secondavenuesagas.com Benjamin Kabak

      Just because you don’t like the results doesn’t mean the projection system is worthless.

      • Dan

        Mo’s Projection means the projection system is probably worthless.

        • http://www.secondavenuesagas.com Benjamin Kabak

          You disagree with Mo’s projection. Either:

          1. You don’t understand how PECOTA, on the whole, works.

          2. The projection system is bad.

          I know what I’m picking here.

  • Brian

    Based on Mo’s projections, they have absolutely no credibility. Absolutely laughable.

    • Accent Shallow

      Mariano is pretty much the definition of an outlier. Any projection system that would nail him would likely overproject anyone else.

      And at some point, he does have to slow down, right? Or is he just going to retire after putting up another 1.50 ERA?

    • whozat

      You’re doing it wrong. Misprojecting one of the biggest outliers of all time in terms of overall performance and ESPECIALLY performance relative to age does not say anything about the utility of a projection system. Systematic error does. For example, we can say that PECOTA is bad at projecting very young players with little big-league track record. That’s true. It may be that, due to the amount of regression they apply across the board, they are bad a projecting the performance of elite players in their prime.

      But rejecting an entire model because it’s wrong on _one_ player, especially one that we know essentially defines the term “outlier” bespeaks a fundamental lack of understanding.

  • http://janeheller.mlblogs.com/yank_c.jpg T-Dubs

    so who is the Yanks pitcher LEAST likely to surpass his projection?

    Burnett if he gets hurt or Joba/Dr. Phil if one of them spends all year in the bullpen I guess.

  • camilo Gerardo

    I want Bill James’ projections!!

  • Mike Pop

    Burnett is going for 16 wins and a 4.22 ERA with 195 strikeouts in 184 innings. Just because that’s all he wants.

  • MikeD

    The problem with projection systems like PECOTA is that any player that has a break-out season will be projected to regress the following year, and any player in his 30s will be projected to also regress. Couple thaht with systems like PECOTA can’t really project young players all that well and suddenly your team seems to suck!

    A few examples that PECOTA projected for 2009

    Pettitte: 10-9, 168 IP, 4.41 ERA.
    Joba: 9-6, 124 IP, 3.09 ERA
    Matsui: .275, 10 HR, 52 RBI
    Jeter: .288, .353, .383
    Cano: .284, .323, .419
    Damon: .280,.354, .423 (12 HRs)
    Posada: .268, .364, .411
    A-Rod: .282, .373, .508, 30 HRs, 98 RBIs (well they nailed this one!)

    Of course, it took a new career-ending situation and a missed five weeks for them to get A-Rod right. They missed on many of the others because the system has to downgrade guys like Pettitte, Matsui, Jeter and Posada because of age. It never tries to calculate come-back seasons, or factor in that Jeter had an off year in 2008 because of some hand injuries.

    This is not meant as a knock of these systems. They’re interesting for trends, but don’t put too much faith in them when it comes to individual players. They’ll get some right and quite a few wrong.

    • bexarama

      Speaking of not being able to predict young players, was it PECOTA that projected Matt Wieters would be the best hitter in the majors last year?

      • Accent Shallow

        Oh, I’m sorry, did you ask me to link this article again?

        /one of my favorites

    • Dalelama

      The new stadium threw them off.

  • jramey

    um with this team that PECOTA gave us, how is that winning the division? horseshit, thats all i have to say.

  • WizardLizard

    So this book probably went to print a bit ago. Are the projections in it the same as the version that projected the yanks to come in third? I’m talking like if you add up the wins and losses of the pitching staff does that total equal the total that placed the yanks third?

  • Rob c

    Good AJ vs Bad AJ…..

    Has anything been said regarding if Bad AJ always tips his pitches as he did in the play-offs vs Angels?

  • DP

    This sounds perfectly accurate to me. In fact, a little optimistic. Every pitcher on the Yankees out performed his talent level last year and frankly the Yankees suck.


  • Lanny

    So now the #’s don’t work out and its time to bash the #’s and say it is worthless??? Glad some of you can be consistent on the #’s

    This is why these projections just be looked at with a huge grain of salt and another reminder that games arent played on spreadsheets.

    • AndrewYF

      The numbers don’t work because the methods used are clearly wrong. When statistical analysis contains significant and obvious errors, then yes, it’s fairly worthless.

      The game isn’t played on spreadsheets, but spreadsheets can predict with fair accuracy what’s going to happen, given past performance and historical trends. But if the spreadsheets are wrong, like BP’s are, then it’s akin to the game being played and runs not being counted arbitrarily. That’s what BP’s spreadsheets are projecting, currently, and that’s why everyone is ignoring them. Man, do they miss Nate Silver.