Projecting Andy Pettitte

Sorting out the rotation, now and later
Open Thread: 3/19 Camp Non-Notes

As exciting as it was to hear the news about Andy Pettitte‘s return, it’s very fair to wonder just what exactly he’ll be able to offer the Yankees once he’s ready to go.  We won’t know the answer to that until he actually gets on a mound in real game situations, but Dan Syzmborski put his ZiPS system to work in the meantime (Insider req’d). ZiPS projects a 4.45 ERA and 1.5 WAR for Pettitte in 125.1 IP this year, accounting for the fact that he missed a full year but was above average before retiring. That’s basically a league average hurler, a bargain at the $2.5M price. For a seventh starter, that’s not bad at all.

Standard disclaimer: Projections are not predictions, just an estimate of current talent level.

Sorting out the rotation, now and later
Open Thread: 3/19 Camp Non-Notes
  • Gonzo

    Sorry, don’t mean to sound like an a55, but the whole projection vs. prediction thing is funny as hell to me.

    • mustang

      Second that.

      • Mike HC

        third that

    • Sayid J.


      According to Google…

      Project: Estimate or forecast on the basis of present trends
      Predict: Say or estimate that a specified thing will happen in the future

      • mustang

        Ok, you’re not helping.

    • mustang

      Maybe when it stops being funny to us we can leave the dark side of the force.

      • Urban

        I predict that we will be projecting Andy’s numbers for the next six weeks.

        • Mike HC

          I think you have that backwards. You project that fans will be predicting Pettitte’s numbers for the next six weeks.

        • mustang

          I wonder if there is a good bar in the dark side because I need a drink after this.

  • Jesse

    I’d sign for that in a heartbeat.

  • TomH

    Standard disclaimer: Projections are not predictions, just an estimate of current talent level.

    What is the statistical performance record of these estimates? I.e., how close do they typically come to what turned out to be the actual performance?

    • Will (the other one)

      Sounds like you’ve got an idea for an Ohlendorfian research project.

      • Sayid J.

        Good god. I go to Princeton as well and that would be one hell of a thesis. Should have suggested it to some of the statistics folks.

    • thenamestsam

      There are numerous other threads on Tango’s blog devoted to looking at the performance of projections in a number of other ways. If you spend time digging around on there they shouldn’t be too hard to find.

  • carlsonjok

    What does the ZiPS system show as a projection for Freddy Garcia?

  • funcrusher

    sure those are reasonable numbers but are they worth replacing hughes’ upside or nova’s projected above average numbers ? i beg to differ (yes to replacing garcia as far as im concerned)

    • Havok9120

      We aren’t talking about doing either of those things yet. We’ll get to see what those two ACTUALLY LOOK LIKE before the team makes any decisions. We’ll even get to see what AP looks like. No need to go off projections or 10 IP in spring training. We’ll have a month or two of the season behind us.

      • Mattchu12


    • thenamestsam

      I agree with you. 1.5 WAR for 2.5M is a bargain. However that 1.5 WAR is wins above a theoretical replacement player. That is not who Pettitte will be replacing. The Yankees are loaded with depth at SP, and even if things don’t go according to plan he’s more likely to be taking the place of Garcia or Warren/Phelps than your prototypical replacement level guy. He’s going to represent at best a marginal upgrade if he truly pitches like a 4.5 ERA guy.

      That said I do think there’s a decent chance he can beat this projection. He’s a pretty rare case as someone who by choice took a year off when he was still a very effective major league pitcher. That makes forming any projection of him quite difficult. Andy always seemed like a guy who would age extremely well to me, and assuming not too much has been lost in the year away I think he’ll be a lot closer to the guy he was before than this projection.

      • Tom

        DEAD ON!

        I think folks need to be careful with the Fangraphs WAR per dollar game. 1 win is also not worth 4.5mil…. 1 win on the FREE AGENT market COSTS 4.5mil. The value of a win is less than 4.5mil. People will say this is semantics, but you can’t confuse price (or cost) on the free agent market with generic value.

        These $/WAR assessments should be considered “did the team pay a fair market price for a free agent?”; especially since wins can be had by other means than signing FA’s. And as the comment above talks about one team’s “replacement” level could be another teams 4th or 5th starter.

        The Yankees could go out and sign a 2.5 WAR CF for 4 mil and it would be a very good deal economically… is it a good move for the team?

        At this point the question should probably be:
        – Is Pettitte a 2.5mil UPGRADE over the current rotation options? I think he can be and with the minor league terms the Yankees have some time to assess this at minimal risk

        (You then also can factor any value in terms of general injury risk in the rotation and any potential value trading one of the other guys as well as any prorated salary saved if they are dealt)

        • King of Fruitless Hypotheticals

          AP won’t be replacing a theoretical replacement player. He’ll be replacing somebody with a torn groin, low velocity, loss of command, etc. He’ll actually be replacing somebody, *IF* that time comes, who is less than replacement and needs to be excised from the rotation.

          • Tom

            And at that time the Yankess have 3-4 options to replace that pitcher(s) with

            He’ll be effectively replacing whoever the other guy the Yankees would have put in the rotation in that scenario (Garcia? Phelps? Micthell? Warren?)

            These aren’t 0 WAR players. Not exactly 3-4 WAR pitchers but not 0 either.

            I like the move as Andy can be better than any of those options and the Yankees are taking on near zero risk to find out if he can be, but it can’t (or I should say shouldn’t) be looked at as a simple above replacement type analysis

  • Havok9120

    I know nothing and can’t help you, but if no one else helps you out here, just repost this in the Open Thread tonight. All the peoples that can help will probably be there.

    • Havok9120

      Hmmm. Apparently the comment I replied to got deleted. OR IT NEVER EXISTED….

      We may never know.

      • Urban

        I do not exist…any more.

  • JoeyA

    “Historical performance is not indicative of future results.”

    This is the lowest of risks (minor league deal $2.5M) with the highest of rewards (#3 pitcher and winningest pitcher in postseason history.)

    Here’s a question, one that would obviously be a result of the Yankees having a great season and great luck with their SP’s, but IF the Yankees make it to the post-season, would they keep Andy out of the rotation in favor of Pineda, Kuroda, & Nova, if all involved have similar, above average seasons.

  • Accent Shallow

    SG over at RLYW ran these, as well.

    Link under my name.

  • Karl Krawfid

    Ima go with a 3.55 era

  • Karl Krawfid

    Yankees could have 2 lefties in the rotation and 2 lefties in the pen. This is like Cashman’s dream fulfilled.

  • RetroRob

    I appreciate the effort, but trying to project what a 40-year-old lefty pitcher, who took a year off, but when last seen was still pitching at near peak levels is all but impossible. The sample size to build the projections has little meaning. The aggregate is pointless. It’s down now to the invdividual, a man named Pettitte.