Losing Derek Holland and a Simple and Critical Truth

In a sense, the pitching market is still at a standstill. Masahiro Tanaka has little reason to sign before his deadline, and other pitchers have little reason to sign before Tanaka does, so for about another week and a half, we could be dealing with a whole lot of nothing. And it’s not just free-agent pitchers. There could be renewed runs at David Price and Jeff Samardzija, but only after Tanaka goes. And free-agent position players might wait for financial clarity as well. The short-term result of all this is that we’re not seeing changes in the 2014 projected standings, because teams aren’t making moves. But something did happen to shake things up right before the weekend.

That something is that Derek Holland hurt himself on his own staircase. It sounds too absurd to take real seriously, but the fact of the matter is that Holland needed knee surgery and he’s projected to be out until midseason. To hear Holland talk, he’s determined to beat those projections back to the field, but medical timetables aren’t made up out of thin air. Holland is going to be out for a significant amount of time. Holland is good, and his replacement will be worse. The Rangers are in a fragile position, and this is a bigger deal than it might seem like.

Holland, last year, was an excellent pitcher who made every start. To the Rangers, he was worth something between four and five wins. Before he got hurt, it would’ve been wise to project him to be worth something between three and four wins. I know this bypasses the details and gets right to the heart of it, but all the details are unnecessary. There would’ve been no reason to think Holland was due to get way better or way worse. There would’ve been good reason to assume a bit of regression, in performance and playing time. I don’t need to explain these things to you — around these parts, this is the fundamental approach.

Now, figure Holland’s out for half the year. That’s something in the neighborhood of 90-100 innings, innings that could now go to guys like Nick Tepesch, Colby Lewis, and Michael Kirkman. That’s not a disaster, but it is a downgrade, because it basically has to be. Those arms aren’t thought to be replacement-level, but it’s easy to envision a loss of about a win or so. That’s on mathematical average, and in reality the season plays out just once, but decisions have to be made based on mathematical averages. Everything — everything — is about the odds.

The Rangers play in a competitive division, and both the Angels and the A’s project to be strong teams in the season to come. The Mariners project to be worse, but they should be at least okay, and they’re probably not done upgrading. Additionally, the American League has its various Wild Card contenders behind the Red Sox and Tigers. It’s going to be difficult to sneak into the postseason, but the Rangers clearly have a quality team.

The new Wild Card rules have existed for two seasons. On average, the Wild Card teams have won 90.5 games, so let’s just knock that down to 90. Let’s set 90 wins as the goal. The Rangers project to be good, and while there’s little sense in identifying a specific number, let’s say the Rangers project for something in the high 80s. Now let’s dock them a win, on account of the Holland injury. Let’s not worry about any issues Holland might have trying to return to full strength and effectiveness. Let’s look at a graph, charting approximate 90+ win odds against a team’s true-talent level.


You’ll recognize this as the win-curve line of thought. The probabilities are approximate, and they ignore the realities of a team’s specific competition, but any given truth will seldom stray far from this. And what’s most important here isn’t the curve itself. Rather, it’s the slope of the curve as the true-talent win total increases.


When you have a really good team, adding a win or losing a win doesn’t change very much about the outlook. If you subtract a win from a team with a true-talent 100-62 record, its odds of winning at least 90 games drop just 1.8 percentage points. When you have a bad team or even a mediocre team, the same idea applies in reverse. Things get volatile when you’re dealing with a team that looks better than .500, but isn’t amazing. A team like the 2014 Texas Rangers.

Between the mid-80s and the mid-90s, a win is worth about 5-6 percentage points. So if you figure the Holland injury costs the Rangers a win on average, that drops their chances of winning at least 90 games by about 5-6 percentage points. And if you’re a bigger Holland fan than that, it doesn’t take much more to get into the double digits. By now this should probably be a familiar argument, because we spend a lot of time talking about the win curve and playoff probabilities, but the Rangers have been hurt because Holland has been hurt. They’re facing lower odds of winning the AL West, and they’re facing lower odds of winning one of the AL Wild Cards. It’s obviously not crippling, but if the offseason is about shifting or fortifying the odds as much as possible, you can think of this as effectively taking a lot of the punch out of the Shin-Soo Choo acquisition. Choo moved the needle forward, by some percentage points. Holland falling on the stairs has moved that needle back, not all of the way, but a significant chunk of the way.

Another way to look at this: while Masahiro Tanaka could be anywhere from great to lousy to hurt, an ordinary projection calls him good, and shy of elite. In other words, an ordinary projection values Tanaka right around the same level where it values Derek Holland. If signing Tanaka would be a splash for the Rangers, this is sort of a half-splash in the opposite direction. Instead of signing Tanaka and leaving fewer innings for the depth guys, the Rangers now have to give more innings to the depth guys, and you never want to have to lean on those.

Which means there ought to be a greater sense of urgency for the Rangers to add, to at least try to offset this to some extent. Jon Daniels has said the Rangers are comfortable with their internal options, but there was also word over the weekend the team is making progress toward signing Jerome Williams. Williams, since returning, has been good for just a 115 FIP- as a starter, but he’s also posted a 101 xFIP-, so it’s not a leap to suggest he could be of some use. What he isn’t is Derek Holland, but he could at least help the Rangers avoid giving too many innings to a mess.

But Holland’s quality is probably irreplaceable, if the Rangers are indeed around their budget limits. Very simply, losing Holland for a few months is bad for the Rangers. Very critically, they’re in a volatile spot, so the injury hurts them particularly hard. The American League just got that much more wide open.

Print This Post

Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.

45 Responses to “Losing Derek Holland and a Simple and Critical Truth”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Juicy-Bones Phil says:

    This article is absurd. Supplying these fantastic win curves means very little without solid data to fill them with:

    “Let’s set 90 wins as the goal. The Rangers project to be good, and while there’s little sense in identifying a specific number,…”

    There is a great deal of sense in identifying a specific number. As an economics major myself, I would never create any sort of model to show a specific goal without identifying very specific outlying numbers. Where are the ZiPS projections that Jeff and Dave normally love to shove down my throat? Where is the mention of a healthy Matt Harrison adding value to the 2014 Rangers? What about the projections for the Rangers internal options vs. the projections for Jerome Williams? Lazy at best.

    I agree that losing Holland is significant, more significant than most people do. I would never try to convince someone of this without a solid base of data to prove it.

    Step it up Jeff.

    -54 Vote -1 Vote +1

    • AMB says:

      The Rangers project for 42 WAR, which if memory serves is a projection of 90 wins-


      +11 Vote -1 Vote +1

    • Yan Fucking Gomes says:

      “I don’t need to explain these things to you”

      Step it up, Juicy-Bones Phil.

      +22 Vote -1 Vote +1

    • Here are numbers:


      Here are numbers:


      My assumption is that FG readers know where those are, since we’ve talked about them a bunch. My assumption is that FG readers don’t need to be explained to at length why the Rangers project to be pretty good. And you’ll notice that the numbers change very little between the mid-80s and the mid-90s, in terms of slope, so it doesn’t matter much where in there you put the Rangers, since the result for our purposes is more or less the same. They’re somewhere within a volatile range, is the point.

      Writing is a balance between providing details and providing readability. Providing more details would hardly have strengthened this piece, considering how much FG readers already know. And every single number makes an article more difficult to wade through. I’d like to write with fewer numbers, not more.

      +65 Vote -1 Vote +1

      • tz says:

        Lazy should not be applied to someone who cranks out an informative article every 3 or 4 hours.

        And just because the analysis isn’t laid out in hyperdetail doesn’t necessarily detract from its validity. Sort of like seeing the phrase “Derek Holland hurt himself on his staircase” and complaining about not providing a GIF of the event to accompany that assertion.

        +39 Vote -1 Vote +1

        • Jeff Who Lives At The Office says:

          I agree with your point, but c’mon. This is a Jeff Sullivan article. We do need him to create a gif of Holland being pushed down the stairs by his dog.

          +12 Vote -1 Vote +1

        • Lou Albania says:

          Could the Notgraphs research team get on that, the gif. Thanks.

          +6 Vote -1 Vote +1

      • John Dupont says:

        I find the original comment valid and Jeff’s response informative and to the point. What I don’t understand is everyone down voting someone who makes a request for more detail…..

        Vote -1 Vote +1

        • Erik says:

          This site could turn every article into a research piece if that’s what the readers really want. The result would be considerably less articles.

          I found this article to be straight to the point, although it could’ve been three times as long for the purpose of specifying whether “high 80’s” meant 86,87,88 or 89. Not sure exactly how much more informed we would actually be.

          Vote -1 Vote +1

      • That Guy says:

        Ha-ha. Burn.

        Vote -1 Vote +1

    • Basil Ganglia says:

      “Where are the ZiPS projections that Jeff and Dave normally love to shove down my throat? Where is the mention of a healthy Matt Harrison adding value to the 2014 Rangers? What about the projections for the Rangers internal options vs. the projections for Jerome Williams? Lazy at best.”

      As an economics major do you start every one of class papers with a review of Econ 101/102 basic micro or macro? Or do you assume that there is a shared body of knowledge and reference data with which economics major are familiar and are capable of using in their work.

      Seems to me that you are the one who is lazy.

      +23 Vote -1 Vote +1

      • Juicy-Bones Phil says:

        There definitely should be a shared body of knowledge as you say, but I would never write a paper on the GDP change of the U.S. before and after the use of interchangeable parts without first stating what the GDP was before and what it became after. I wouldn’t ever assume, and neither should anyone else.

        -17 Vote -1 Vote +1

    • Luke says:

      Nobody cares what your major is, bro.

      +17 Vote -1 Vote +1

    • Catoblepas says:

      Negative Vote Farming
      Step 1 — insult Jeff Sullivan after a solid, interesting article
      Step 2 — success!

      Vote -1 Vote +1

    • cs3 says:

      daaaaamnnnn Juicy Bones Phil just got bitch slapped!

      Vote -1 Vote +1

  2. Juicy-Bones Phil says:

    I don’t want links or assumptions, I want to see the data in the article.

    -39 Vote -1 Vote +1

    • TK says:

      I know and how dare they assume we know what “ERA” is or a “win” or who “Derek Holland” is or what a “Texas Ranger” is. The more extraneous information the better! All areticles must be 50 paragraphs of repetition for every 2 of content.

      +19 Vote -1 Vote +1

      • BoKnows says:

        An “era” is a time period of distinction, a “win” is a hipster term for window, “derek holland” is the full proper name of a medieval country, and a “texas ranger” is a law man made famous by our lord and savior Chuck Norris.

        +36 Vote -1 Vote +1

        • Go Nats says:

          I voted thumbs up just for mentioning Chuck Norris with the Texas rangers. After all he did count to infinity twice!

          Vote -1 Vote +1

      • gnomez says:

        What’s a baseball?

        Vote -1 Vote +1

  3. Chicago Mark says:

    Way to turn a good article into a discussion about yourself juicy. Giving more time to you is a waste. So onward and upward. Good job Sully. I like your bio the best. I believe most of the time. But not all of the time. I think.
    Damn, I have Holland for a buck as a holdover. Damn! Damn!!

    Vote -1 Vote +1

  4. Yan Fucking Gomes says:

    Your critique is lazy at best.

    Vote -1 Vote +1

  5. Lou Albania says:

    I want a ham sandwich, with pickles.

    Vote -1 Vote +1

    • I. P. Freely says:

      The fact that this article did not include a ham sandwich with pickles has me really really pissed off.

      Vote -1 Vote +1

  6. RC says:

    I’d argue that its a little ridiculous to think that an additional 1WAR in the 1st year of a big 100+M contract should have any affect on the thought process of the Rangers is a bit absurd.

    This shouldn’t change how they view Tanaka one little bit.

    Vote -1 Vote +1

    • murphym45 says:

      It shouldn’t affect their view of Tanaka, but it is a significant injury, so why wouldn’t it affect their thought process? Granted, that thought process is more likely to be about how to bring in a decent, inexpensive rotation depth piece, not whether to sign one of the top remaining free agents to an expensive multi-year contract. Still, it’s a significant injury in a volatile division, so it would be a little ridiculous to think that it wouldn’t affect their thought process in some way.

      Vote -1 Vote +1

    • Vivek says:

      Who is saying it changes the view of Tanaka? I think what it does is change whether they want to sign him this year. A team with 3 good starters would probably want Tanaka more than a team with 5 good starters. The Rangers want to contend this year. If this was the Cubs, yeah, an SP injury probably doesn’t change much, Tanaka is a long-term play.

      Vote -1 Vote +1

  7. Dwayne Carter says:

    it went downhill quick upton here, upton here.

    i would give thumbs up to the comment about the bros major, but my thumbs never seem to show up, plus i be on grad students so much, i’m talkin T.A., and unlike the fed, ain’t scared to inflate.

    Vote -1 Vote +1

  8. Daniel says:

    While I agree with most of the points in the article, I have a little trouble with the statement, “A team that looks better than .500, but isn’t amazing. A team like the 2014 Texas Rangers.” While they certainly aren’t a clear-cut top 3-5 team, describing a team as “better than .500” feels a bit like the Cleveland Indians or something.
    Per the projected standings page, the Rangers rank 5th in the majors in projected W/L record.

    I’m certainly not agreeing with Mr. Juicy-Bones, but I think that the Rangers aren’t exactly a team that projects for upper 80s wins. The W/L projections page has them at 85, but then again they’re still fifth and no team is higher than 89.
    With the return of (overrated) Matt Harrison and the acquisitions of Prince Fielder and Choo, it’s hard to imagine the Rangers not improving on a disappointing 91-win 2013.

    Vote -1 Vote +1

    • Brad Johnson says:

      It’s worth pointing out that the projected standings pages should not be taken literally. They’re based on an imperfect projection system and educated guesses about playing time. In essence, they’re a tool that supply trivial knowledge. The kind of knowledge that contains enough of a fudge factor that it’s exceedingly pointless for Jeff to determine if the Rangers are an 86 or 89 win true talent team. He was right not to do so.

      Vote -1 Vote +1

  9. Steve says:

    My troll posts are way better, kid. Step it up (I don’t troll all the time, I promise).

    Vote -1 Vote +1

  10. AC of DC says:

    Did they really go with a “tripped over his dog on the stairs” story? Isn’t that one of the most overused b.s. covers for an athlete’s injury in American sport? Can we get some numbers on how many guys have “tripped over dog on stairs” on a graph against guys who “slept on it wrong”?

    Vote -1 Vote +1

  11. Dustin Dietz says:

    Great stuff. While losing Holland for a period of time will hurt the Rangers, I think one has to also consider the other teams will lose key players to injury for periods of time as well. One cannot predict injuries, but a team like the Tigers that had all five rotation members rarely miss starts, should not expect such outstanding health again this year.

    Vote -1 Vote +1

    • Forrest Gumption says:

      The current Rangers rotation is Darvish-Harrison-Ogando-Perez-Lewis. Harrison and Lewis are coming off injuries, Ogando had a 4.64 xFIP and Perez is still too young to project, while never dominating. Darvish is a true ace.

      The current A’s rotation is Gray-Parker-Kazmir-Straily-Griffin with Milone, starter of 57 games the last two seasons, at AAA as insurance. Gray was absolutely dominant in his 10 starts and throughout the playoffs, Parker pitched while injured and threw a month of terrible games but then came back strong, Kazmir had really amazing peripherals, Straily and Griffin weren’t dominant but are probably #3s on 75% of teams.

      Up until the Holland injury both teams were fairly even, but losing 15 starts from Holland and replacing them with Jerome Williams is massive and definitely pushes the A’s as the #1 seed in the West.

      Vote -1 Vote +1

      • Brad Johnson says:

        As I concluded in my own article on this topic, the Rangers better hope their offense is scary good.

        Vote -1 Vote +1

      • algionfriddo says:

        I have read several times that Robbie Ross was going to get another shot at a rotation spot & this was before the Holland injury. Ross actually pitched well in spring training as a starter in 2013 but had an uneven year vLHB during the regular season out of the pen. I hope they give Ross a true look but he has been very useful in the pen. Any Rangers fans heard of an update on Ross? Since Ross is a lefty and they already have several lefty starting options they may decide to keep him where he has been… in the pen, where they don’t have as many lefty options.

        Vote -1 Vote +1

  12. PackBob says:

    In an economics class, the reason to cross all the t’s and dot all the i’s when presenting a model is so that the professor knows the students know what they are doing and to insure the students are using the right processes. One could also demand citations for any saber metric principles employed and thanks to Jeff’s parents for supporting his work.

    Vote -1 Vote +1

  13. Wobatus says:

    The myth of the 5 man rotation keeps asplodin’.

    Vote -1 Vote +1

  14. Juan says:

    Mr. Sullivan fails to consider the impact the upgraded
    offense will have on the rest of the pitching staff.
    Additionally, the new young pitchers from last year
    (i.e. Perez, Tepesch, Cotts, Ross etc.) are a year
    older and a year more experienced. And, there are more
    arms in the minors.

    Of course, Mr. Sullivan could have made his article
    more realistic by doing like an Economist and starting
    the article by stating, let’s “ASSUME” all other
    things remain equal …

    Vote -1 Vote +1