The Evolution of Groundball Rates

On Wednesday, Dave Cameron ended his piece entitled “Halladay and Verlander” with this note, seemingly fired at those of us in the prospect business:

It might not be as sexy, but getting groundball outs and limiting walks is a far more efficient way of pitching than trying to blow every hitter away. Strikeout rate is nice, but don’t let it be the only tool you use to evaluate a young pitcher – not only are Ks not the only way to succeed, they aren’t even the best way.

This is a belief system that isn’t new to me, as it was essentially the premise for my offseason Staring Down the Sinkerballers series. But, that piece was limited to discussing one type of player: the two-seamer/slider pitcher that often flies under the radar. The fact is, we need to do a better job understanding how all pitchers will see their batted balls translate from minor league baseball to the Bigs. This is a case study of just that.

I created a list of pitchers that debuted in 2009, have had 150 innings over the span of two years, and at least 150 batters faced in each season. If I did my search at the B-Ref Play Index right, and I whittle it down to players with minor league experience, there are 18 pitchers left. It’s a diverse list, with lefties and righties, groundballers and flyballers, good pitchers and bad. It’s also a group that had their entire minor league career in the MinorLeagueSplits era. Now, I’ll tell you this: there are flaws with the mLS data. We all know that. But, like anything else, until we have something better, I’m going with that. I’ll probably follow this post up at a later time using GB/FB ratios — which I have official data for — but that seems just as flawed.

So, I looked at all 18 players, and I compiled a list at the levels where they faced 150 batters (this number I keep using is important: it’s when groundball rate stabilizes for a pitcher, at least at the big league level). Let’s start, both as an introduction to my sample group and as an entryway into the data, with the career minor league (mLGB) and Major League groundball rates of each player. I have ranked them in the order of those that dropped most from the minors to Majors:

Name                MLGB      mLGB      Diff
Brett Cecil         43.2      59.7      -16.5
Brian Matusz        34.7      48.1      -13.4
Rick Porcello       51.6      63.7      -12.1
Vin Mazzaro         41.9      52.8      -10.9
David Hernandez     28.5      37.9      -9.4
Trevor Cahill       51.2      58.6      -7.4
Jason Berken        41.3      48.6      -7.3
David Huff          37.2      44.3      -7.1
Craig Stammen       48.4      54.1      -5.7
Wade Davis          40.5      45.7      -5.2
Kris Medlen         42.2      47.2      -5.0
Mat Latos           43.1      47.7      -4.6
Brad Bergesen       49.0      52.6      -3.6
Brett Anderson      52.4      55.5      -3.1
Brian Duensing      48.5      50.3      -1.8
Doug Fister         47.8      47.9      -0.1
Tommy Hanson        40.2      39.1      +1.1
Ricky Romero        54.0      48.6      +5.4

It’s interesting, of course, that a different Blue Jays lefty is on both the top and bottom of this list. But also interesting that half the players dropped their groundball rate between 3 and 7.5 percent between the minor and Major leagues. Those who attended FanGraphs Live heard the problems I have with using MLEs to project minor leaguers. While it would be easy to do this study with everyone in the MinorLeagueSplits era, create an average drop that we apply for everyone, I just don’t find it all that informative. If you want to assume a player drops about 5-5.5 percent when he reaches the Majors, you’ll probably be about right as often as you’re about wrong.

But, let’s take it a step further. I next brought the 18 into four quadrants, ranked by their MLB GB%. Here are the five least groundball-ing pitchers, and their groundball rates at each level they registered 150 batters faced at (Note: The ~ symbol appears when they did so at that level in multiple seasons. I’m eyeballing there. Again, there are error bars around this data that are unavoidable).

Name       Low-A   HighA   DoubA   TripA   MLB
Hernandez  36.8    37.6    36.7    38.8    28.5
Matusz      N/A    49.5    45.8     N/A    34.7
Huff        N/A    42.0    49.5    ~44.0   37.2
Hanson     41.6    38.6    40.7    36.5    40.2
Davis      49.5    47.5    ~46.5   ~39.6   40.5

I find this far more explanatory. If we broke this group into guys with similar stuff profiles, you’d certainly group Hanson and Davis together as hard-throwing right-handers, with similar velocities on their fastballs, sliders and curveballs. Hanson was always the flyball pitcher he’s proven to be, but Davis took some trending down before showing his true colors in Triple-A. Hernandez is close to that caliber, but his stuff was never viewed in the same light. Like Hanson, he was always a flyball pitcher, but has taken it to a new level in the Major Leagues. He’s ditched his slider, and has problematic command, which I think might be to blame for the sub-30 GB%.

I’d also group Matusz and Huff together — college lefties with good command that start their arsenal with a 90 mph fastball and an 82 mph changeup. Neither had dependable Double-A numbers in the slightest, and with their stuff, shouldn’t have been trusted as 45% guys. I could see looking at all similarly-profiled lefties and searching for something similar. But for now, we look at our next quadrant, the flyball-leaning pitchers:

Name       Low-A   HighA   DoubA   TripA   MLB
Berken      N/A    52.0    47.6     N/A    41.3
Mazzaro    56.4    53.9    52.7    ~49.2   41.9
Medlen      N/A     N/A    43.4     N/A    42.2
Latos       N/A     N/A    41.3     N/A    43.1
Cecil       N/A     N/A    59.2    59.8    43.2

Wait, there’s another college lefty with good command, a fastball around 90, and a changeup at 82, also posting some misleading Double-A numbers! Brett Cecil, come on down! Cecil throws a sinker more often than Matusz and (especially) Huff, and it really worked in the minors. Whether it’s an issue of command, or just not enough sink, it hasn’t in the bigs. Latos probably fits in with Tommy Hanson, never really hiding that he’d be a flyball guy. Medlen is just outside of that category, smaller than those pitchers, with less velocity, and a change-up as his go-to pitch. In a small sample, he was getting a ton of groundballs in the low minors, but in Double-A, showed the tendencies he’s settled into with Atlanta.

Finally, I think we have to compare Jason Berken and Vin Mazzaro, similarly sized fastball-slider guys. Mazzaro throws more sinkers, and thus, has always had higher groundball rates. But ultimately, I think this is a sign of the quality of Major League hitters: neither pitcher has stuff that’s considered particularly world-beating. Next, we have four guys that have induced better-than-average groundball rates:

Name       Low-A   HighA   DoubA   TripA   MLB
Fister      N/A     N/A    ~48.0   44.1    47.8
Stammen    55.9    ~53.5    N/A    ~54.0   48.4
Duensing    N/A    48.8    ~48.8   ~50.2   48.5
Bergesen   ~48.5   54.7    53.1     N/A    49.0

Having introduced Mazzaro and Berken, smallish sinkerballers, gives us a nice frame of reference with Brad Bergesen and Craig Stammen. Both of them are of a similar ilk, though maybe Stammen’s cutter separates him a bit, and maybe explains his greater success than Mazzaro in maintaining an above-average GB%. Bergesen is explained, I think, simply by having a better sinker. But the fact is, with all four, we do see a noticeable drop from even Double-A to the Majors. Fister is a bit of a weird duck, and hard to explain — you have to think his size, more than anything else, leads to a good groundball rate. But that’s sort of always been the case, I guess. Part of you wants to compare Duensing to the other college lefties above, but he throws a little harder, has spent most of his Major League time in relief, and his go-to offspeed pitch is a slider. And he hasn’t wavered much in terms of groundball rates.

Finally, here are the guys getting a ton of groundballs:

Name       Low-A   HighA   DoubA   TripA   MLB
Cahill     56.4    61.4    61.8     N/A    51.2
Porcello    N/A    64.1     N/A     N/A    51.6
Anderson   57.9    ~54.8    N/A     N/A    52.4
Romero      N/A    41.9    ~48.7   55.9    54.0

Ricky Romero is obviously the guy that stands out, but I hesitate to make many assumptions from his numbers. It’s pretty clear that he didn’t become interested in inducing grounders until 2008. This probably sounds like a cop-out, but it’s true: go back and read Baseball America’s scouting report of Romero from the 2005 draft — the movement on his fastball isn’t lauded at all. Fast forward three years, and they are praising the “good life” on his fastball, and referencing a new “vulcan change, which behaves like a splitter.” He’s just a different guy.

I would not hesitate to compare Cahill and Porcello, two lauded high school pitchers that are big, throw sinkers 50% of the time, and were 60% groundball guys in the minors. They are ridiculously similar, and I think tell us a lot about their type of pitcher. Finally, we have Cahill’s teammate Brett Anderson, who might be too unique to comp. His combination of movement, velocity and offspeed stuff is pretty unmatched among Major League lefties. It should be no surprise that the truly unique can sustain +50% ratios at the highest level.

That’s where we’ll leave things today. But there are some interesting groups that I think we can extrapolate and further investigate: the fastball-change lefty, the prospect-y hard-throwing righty, the smallish sinkerballer, among others. If we can understand players in the context of their own stuff, then we might be getting to a place where we can evaluate young pitchers using the tool that Mr. Cameron hopes.

Print This Post

11 Responses to “The Evolution of Groundball Rates”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Brandon T says:

    I think the increase emphasis on the groundball is as much a consequence of the “Steroid Age” as anything else. When hitters evolve towards lumbering behemoths that want to hit a mile in the air, the natural way to counteract this is to induce groundballs, which of course reduces slugging. This might be what has, in part, led to “the Year of the Pitcher” which some people have been hastily proclaiming. I think it would be really interesting to look at groundball rates vs. SLG for all of MLB, to see if there is a connection. Another aspect of this is that, until recently, sinkerballers would undervalued, as they tend to underperform in the minor leagues due to worse defenses and field conditions. Basically, as a young pitcher, you are either directly or indirectly encouraged to strike people out — that’s what gets rewarded in most cases. The guys who get the big bonuses and fast promotions are those who strike people out. It’s fun to look at the minor league performances of a lot of top sinkerballers — it tends to improve as they move up the ladder. It’s less of a issue now, though, as it has become known as a valuable and repeatable skill.

    Vote -1 Vote +1

    • joser says:

      But isn’t it rather late in the “Steroid Era” for something like this to show up? In baseball-career terms, we’re well past a generation into it. Even if the scouts and pitching coaches in the lower levels didn’t change their emphasis until Barry Bonds’ head started casting shadows all the way to the East Bay, you’d have expected to see this show up before now.

      I think the rest of your argument is much stronger — groundballers get punished by poorer defenses at the lower levels, and K rates are the traditional stat that everyone (over) values.

      Vote -1 Vote +1

  2. slash12 says:

    This is an excellent article, I’ve often pondered the transition of ground ball rates from the minors to the majors, and noticed that they always tend to trend downwards at varying amounts.

    Vote -1 Vote +1

    • slash12 says:

      Another interesting study in this area, would be to analyze walk rates, and strikeouts in the same fashion. There are some guys with extremely high strikeout rates in the minors, that can’t seem to translate that at the major league level, while others can maintain the high rate almost instantly.

      Vote -1 Vote +1

      • Bryan Smith says:

        I agree, and I’ll put that on the to-do list for this winter. I’d go back to 2002, as far back as FG has pitch type data for.

        Vote -1 Vote +1

    • JMHawkins says:

      Most stats do. The hitters are better, after all. If you make a mistake and leave a fastball up against a AA hitter, he maybe fouls it off and you get him to ground out on the next pitch. Or else, impatient youngster that he is, he chases stuff and gets himself out with a weak grounder. Move up to the majors and the ball up in the zone lands in somebody’s soup at the outfield cafe and guys are better at laying off marginal pitches.

      Vote -1 Vote +1

  3. NadavT says:

    Nice use of “belief system” there!

    Vote -1 Vote +1

  4. Nathaniel Dawson says:

    Harry Pavlidis posted a look at groundball rates through the different levels of the Minor leagues:

    GB rates appear to fall the farther up you go, agreeing with what you saw from the pitchers you looked at.

    Vote -1 Vote +1

  5. Nitram Odarp says:

    I know this may be a little off topic, but I’d love to see something on the insane season Tim Hudson is having in terms of batted balls. I’m not sure we’ve ever seen a better example of a pitcher pitching to contact since I can remember. He’s currently got the third highest GB rate of fangraphs batted ball era, to go along with a LD rate that’s 1.5% less than the next best of the same era at a measly 11.8% (Lowe is next best at 13.3% in 2002). In spite of his less than impressive K rate and BB rate he’s still got a very good tERA (3.48, compare to Halladay at 3.20). I don’t even know what to make of this…it’s just crazy

    Vote -1 Vote +1

    • Ethan says:

      David Golebiewski looked at Hudson and found him to be getting very lucky with his BABIP on ground balls ( tERA adjusts for this, I think, which explains the 1.3 runs it adds to his ERA.

      However, I don’t think tERA adjusts for his low LD%. LD% is almost never repeatable year-to-year, and Hudson has a much higher career LD%, so the best guess is that he’s just getting lucky re line drives this season.

      Assuming more line drives (which also means fewer GBs), you’d have to bump Hudson’s numbers up closer to his FIP or xFIP totals: 3.89 and 4.03 respectively.

      I wrote a stat which does all of these adjustments, and it has Hudson coming in at 3.94 this season:

      Vote -1 Vote +1

  6. Mr Punch says:

    “But the fact is, with all four [better than average GB guys], we do see a noticeable drop from even Double-A to the Majors.” Well, actually, no we don’t. Two out of four, maybe.

    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>

Current day month ye@r *