Mile Fly City?

Recently, one of our readers, Simon, noted that the Rockies might be targeting fly-ball pitchers with the recent additions of Guillermo Moscoso, Jamie Moyer and Jeremy Guthrie. I decided to examine if going after fly-ball pitchers was a practical method for limiting runs at Coors Field.

In an ideal world, the Rockies would love to have all extreme sinker-ball pitchers. The Rockies GM, Dan O’Dowd, stated this stance recently on Clubhouse Confidential.

In an ideal world, every single guy in Colorado would be a heavy sinker ball guy who would have a tremendous ground ball to fly ball ratio.

It is not an ideal world and he knows it. He goes on further to state:

Unfortunately not all of our decisions are made in an ideal world. When we balance fly ball rates, we really try to balance soft and hard.

To see if using extreme fly-ball pitchers would be an advantage for the Rockies, I calculated, using Retrosheet data, the batted-ball data for all hitters at Colorado and the rest of the league. I looked at the numbers from 2002 (the first year the humidor was used at Colorado) to 2010 (the last year I have imported the Retrosheet files). Here is the data:

Ground Balls AB BABIP AVG SLG ISO HR% % of Batted Balls
COL 17543 0.214 0.214 0.234 0.021 0.0% 46.3%
League 489744 0.226 0.226 0.246 0.020 0.0% 46.3%
Fly Balls
COL 11919 0.156 0.253 0.706 0.453 11.5% 31.5%
League 363535 0.139 0.221 0.598 0.376 9.6% 34.3%
Line Drives
COL 8433 0.726 0.731 1.006 0.274 1.8% 22.3%
League 205104 0.705 0.712 0.979 0.267 2.3% 19.4%

First off, it is obvious that ground balls are the best end result for a pitcher in Colorado, or anywhere else. The advantage of having a fly-ball pitcher is not so apparent. In Colorado, hitters will have a higher AVG (by about 30 points) and ISO (by about 80 points) on fly balls vice elsewhere. These increases are more than those seen in AVG (increase of about 20 points) and ISO (increase by fewer than 10 points) by line drives in Colorado when compared to the rest of the league. Having a fly ball hit off a pitcher in Colorado is not an ideal outcome.

Colorado states that it is actually looking for pitchers who induce weak fly balls, so I collected a few stats for the pitchers they added this offseason. First, I got the pitcher’s career batted-ball data. Besides the batted ball data, I got the number of hard hit fly balls (> 250 ft) and soft hit (80 ft to 250 ft). I used the batted-ball distances from MLB’s Gameday data (same data that Pitch FX data come from). Ideally, Hit FX data would be used, but it is not available to the general public. Here is the data for the Colorado pitchers. I included the MLB average values and the numbers for Jhoulys Chacin, a ground-ball pitcher for Colorado, and Matt Cain, the poster child for weak contact on fly balls:

Name LD% GB% FB% IFFB% HR/FB 80ft-250ft >250ft Weak FB%
Jeremy Guthrie 18.3% 40.6% 41.0% 10.3% 10.2% 429 623 40.8%
Guillermo Moscoso 17.7% 26.8% 55.5% 12.4% 6.2% 81 89 47.6%
Josh Outman 16.6% 39.6% 43.9% 7.3% 4.9% 62 88 41.3%
Jamie Moyer 14.7% 44.1% 41.1% 8.6% 13.2% 301 392 43.4%
Tyler Chatwood 22.0% 47.0% 30.9% 7.5% 9.6% 58 82 41.4%
Jhoulys Chacin 17.7% 52.4% 29.9% 9.7% 10.7% 71 160 30.7%
Matt Cain 18.5% 37.2% 44.2% 12.6% 6.5% 477 629 43.1%
ALL 18.2% 44.3% 37.5% 9.3% 82700 114393 42.0%

Of the pitchers who Colorado picked up, Moscoso looks to be the ideal model for weakly hit fly balls, with 12.4% IFFB and HR/FB of 6.2% — a weak-hit fly ball rate over 5 points more than the league average. Matt Cain and him have similar batted ball profiles. The knack for inducing weak contact can be seen in the pair’s ability to have have a better career ERA than their FIP, xFIP and SIERA

Moscoso, 3.47, 4.22, 5.05, 4.96
Cain, 3.35, 3.69, 4.26, 4.20

Of the other pitchers brought in this offseason, Jamie Moyer is the closest to fitting the model of a pitcher who induces weak-hit fly balls. His weak hit % is similar to Matt Cain. Both are near 43%. Moyer has seen his ERA outperform his ERA estimators over his career, but not to the level of Moscoso and Cain:

Moyer, 4.24, 4.46, 4.65, 4.66

Jeremy Guthrie and Josh Outman both have above average fly ball rates (41% and 44% vs 38%). Their problem is that they seem to allow long hit fly balls at an above average rate (both at 41% vs 42% for league). Even though they both allow an above average number of hard hit fly balls, they have seen their ERA stay below the other ERA estimators:

Guthrie, 4.19, 4.68, 4.61, 4.50
Outman, 3.75, 4.01, 4.53, 4.51

Also the Rockies traded for Tyler Chatwood. Chatwood will join Jhoulys Chacin as one of the sinker ball pitchers vying for a spot in the rotation. The Rockies acquired two other pitchers, Alex White and Drew Pomeranz, as part of the trade that sent Ubaldo Jimenez to Cleveland. The pair are reported to be ground-ball pitchers. Currently, they look to have little chance of making the initial MLB pitching staff.

Dan O’Dowd was correct that they Rockies did bring in a couple of weak fly-ball pitchers in Moscoso and Moyer. Guthrie and Outman don’t seem to induce weak-hit fly balls, but have been able to outperform their ERA estimators in the past. While it is not ideal to have fly-ball pitchers, especially in Colorado, the Rockies have acquired some pitchers at the extreme end of the fly-ball spectrum. These pitchers allow fewer hard-hit fly balls and therefore induce more outs. There actually seems to be some data to suggest that there is some method to the madness for the wide range of pitchers acquired by Colorado this off season.

Print This Post

Jeff writes for FanGraphs, The Hardball Times and Royals Review, as well as his own website, Baseball Heat Maps with his brother Darrell. In tandem with Bill Petti, he won the 2013 SABR Analytics Research Award for Contemporary Analysis. Follow him on Twitter @jeffwzimmerman.

21 Responses to “Mile Fly City?”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. In Oakland, aren’t most flyballs weak? Nobody has ever been successful in Colorado with a ground ball % as low as Moscoso’s.

    Vote -1 Vote +1

    • Dick says:

      I disagree.

      -18 Vote -1 Vote +1

    • byron says:

      Doesn’t Oakland have enormous foul territory that turns a lot balls that would essentially go unrecorded as foul balls into IFFBs? Also, HR/FB rates aren’t going to do you much good unless they’re adjusted for parks, if not also competition.

      Vote -1 Vote +1

  2. DrBGiantsfan says:

    Interesting study. The problem with Coors Field and other parks at elevation(a lot of PCL AAA parks), is not just that balls fly farther in the thinner air. Since there is less air resistance, it is harder to get movement on pitches.

    While sinkerball pitchers get more grounders, it’s also true that they tend to give up a lot of HR’s if their pitches aren’t sinking and sinkers don’t sink as much in Coors. We all know what happens to breaking balls that don’t break.

    Pitchers are going to give up more runs in Coors than in average parks. The trick is to give up less than other pitchers. My long held theory based on the physics of the situation is that the ideal pitcher for Coors are power pitchers with good FB command. Matt Cain would actually be the prototype. Such a pitcher gives you K’s and a low BABIP. A modestly increased walk rate in return for keeping the ball out of the fat part of the strike zone is acceptable.

    Yeah, a guy like Cainer will give up a dinger now and then but the damage is limited by the low BABIP’s and low HR/FB and that’s all you need to do in Coors, limit the damage.

    Vote -1 Vote +1

    • That’s pretty much exactly the blueprint. GBs are important, but K’s are most important. That’s why Ubaldo was successful and Chacin to a lesser extent.

      Vote -1 Vote +1

    • Jake says:

      The article linked below is a great analysis of the effect of altitude on different pitches. He finds that fastballs are most affected, not breaking balls as many think. This used to be hosted on Athletics Nation, but was relocated by the author.

      Vote -1 Vote +1

      • etrain says:

        The research is correct. The reason breaking balls were viewed as not being effective at Coors is because, pre-humidor, the balls would dry out and make it more difficult for the pitchers to grip the ball effectively enough to throw the off-speed stuff.

        This was reported by many pitchers at Coors pre-humidor; one said that the baseballs at Coors felt like throwing a cue ball. By many reports, one effect of the humidor is that the balls dry out less and therefore the pitchers are able to grip the ball more effectively.

        Vote -1 Vote +1

      • joser says:

        Drier balls also have more elastic collisions, meaning they jump off the bat. That’s going to be most obvious with hits on fastballs, since that’s the most energetic collision and thus the elasticity (which tends to be non-linear wrt forces involved) will have the greatest effect.

        Vote -1 Vote +1

      • DrBGiantsfan says:

        If the pitcher is depending on fastball movement, then I would agree it make physical sense, but a pitcher who is throwing his FB fairly straight but with precise location, the altitude should either be a neutral or slightly positive factor.

        Vote -1 Vote +1

      • Andrew says:

        Rafael Betancourt would be a great example of a guy that thrives off a straight fastball at Coors.

        Vote -1 Vote +1

    • jim says:

      part of me hopes the giants let cain reach FA and the rockies offer him gobs of money

      Vote -1 Vote +1

  3. Don’t forget about Kevin Slowey, whom the Rockies had for a grand total of a few weeks.

    Vote -1 Vote +1

  4. Psst says:

    Pitchers like Jorge Delarosa who saw his GB% jump from 40% in KC to as high as 52% in 2010 in Denver may be the reason the Rockies are optimistic about adding guys like these. Bob Apodaca has had great success working with Jason Hammel, Matt Belisle, Taylor Bucholtz and many others helping them to control BB% and add GB% so out of all of their additions I think Chatwood will benifit the most.

    Vote -1 Vote +1

    • Josh says:

      LOL @ citing Jason Hammel as a model of success

      Vote -1 Vote +1

      • Psst says:

        You mean LOL the guy who went from a 5.25FIP to 350 innings of 3.7FIP? for a 3-4 starter in 2009-10 Hammel was great. He wasn’t an ace but he was certainly worth the C prospect they gave up to get him.

        Vote -1 Vote +1

      • Josh says:

        Ah yes lets just ignore that little thorn of a season in 2011 with the 170 innings of 4.83 FIP ball. Cherry-picking the less relevant data to fit your argument are we now?

        Vote -1 Vote +1

      • Psst says:

        So if Hammel posts two 4(3.9) win seasons followed by a bad one the the first two weren’t a success? Who’s cherry picking now?

        Hammel fell apart last year (although he seemed to recover after moving to the pen,7G 2GS 24 innings 1.85 ERA 17K 5BB) but that doesn’t mean he didn’t improve upon arriving in Denver.

        I would certainly call Joel Pineiro a Dave Duncan success story, regardless of how he pitched last year.

        Vote -1 Vote +1

  5. CJ in Austin TX says:

    I know that DIPS isn’t absolute, but are we just skipping over the intermediate steps of proving the exceptions, so that we can ignore DIPS? Even assuming that there are individual exceptions with respect to weak fly ball rates and regressing HR/fly rates to league average, I would not be comfortable that anything you have done in this article helps us predict those players. Is the 80-250 vs. >250 Flyball tendency a repeatable skill (controlling for the same skill that produces strike outs)? Even beyond that, are the samples adequate to say that these pitchers are definitely short or long fly ball pitchers? Moscoso has less than 150 IP, for example. The data in the article doesn’t provide much evidence that the short vs. long fly ball distinction results in better fly/HR rates. Jamie Moyer has a similar profile to Matt Cain, yet his fly/HR rate is above league average. Guthrie and Outman have below average ratios of short fly balls, and their fly/HR rate is also below league average.

    Vote -1 Vote +1

    • DrBGiantsfan says:

      I’m not sure what profile you are talking about in regard to Jamie Moyer vs Matt Cain. In what way are those two pitchers similar?

      Here’s where you have to take what you see on the field and square it up with the stats. Matt Cain throws hard fastballs up in the zone with good location and he doesn’t give in to hitters. Once you understand that, the statistics make perfect sense and you have a model for how other pitchers can have the same kind of success.

      I haven’t seen enough of Guthrie or Outman to know if they fit the Cain profile, but based on what I’ve heard of Guthrie, I’d guess he’s somewhat similar.

      Vote -1 Vote +1

      • CJ says:

        With respect to Moyer’s “profile,” I’m referring to this quote from the article:

        “Of the other pitchers brought in this offseason, Jamie Moyer is the closest to fitting the model of a pitcher who induces weak-hit fly balls. His weak hit % is similar to Matt Cain. Both are near 43%. Moyer has seen his ERA outperform his ERA estimators over his career, but not to the level of Moscoso and Cain”

        Vote -1 Vote +1

      • DrBGiantsfan says:

        Well, that’s why you can’t just look at a stat sheet and conclude that two pitchers are similar, because Moyer and Cain are about as dissimilar as you can get.

        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>