Felix Hernandez, Groundball Machine
After two starts, Felix Hernandez has taken Seattle by storm. He has a 0.69 ERA, and we’ve run out of adjectives to explain how great he looked on Tuesday. Ron Gardenhire referred to The Royal Curveball as “a curveball from hell”, for example.
Well, there’s one area of Felix’s pitching that I think deserves a little more attention than it has received. His groundball dominance has been nothing short of amazing to watch. According to ESPN’s tracking (via Stats Inc), through two starts, Felix has induced 26 groundballs and allowed 6 flyballs, for a 4.3 to 1 ratio. The Hardball Times, which gets proprietary play-by-play data from Baseball Info Solutions, has his mark at 5.4 to 1. And, according to Baseball Prospectus (which, I believe, is simply measuring groundouts and flyouts), Felix’s mark stands at 5.0 to 1. Due to the differing methodologies by the three sites, finding uniformity is nearly impossible. But, regardless, they’re all generally measuring the same thing, and by any of the three measures, Felix’s G/F rate so far has been remarkable.
I knew he was a dominant groundball guy in the minors as well, so this performance wasn’t a huge surprise. The question that came to me, however, is how well could we accurately determine what his “true” groundball/flyball ratio might be going forward from such a small sample size. Since we know that pitchers have strong consistent abilities to influence G/F rates, we’re fairly certain that Felix’s groundball dominance is a repeatable skill. But, we don’t know if the insane groundball inducing power he’s held over hitters through two starts will continue at this level. It’s almost certain that Felix is going to be one of the more dominant groundball pitchers in the game, but is he going to be a groundballer of Derek Lowe proportion or Brandon Webb proportion? Or will he just create his own new level of groundballing not yet seen, which is what he has done through two starts?
There’s no perfect way to answer this yet, and the small sample size police are going to run over this thread, I’m sure, but my theory is that dominant groundball tendancies in a small sample are indeed evidence of a trend that we can expect going forward. For instance, I don’t expect to ever see Julio Mateo induce 12 groundouts and 0 flyouts in an outing. That kind of performance is just beyond his skillset and the way he pitches. So, what I decided to do was to look through the gamelogs of some major league pitchers and attempt to determine the variance of their G/F ratio in a particular game versus their established G/F rate for the season. If we see that there is little variance from game to game, we can assume that this is a repeatable skill and Felix’s absurd groundballing ways may indeed continue to the point that he becomes the most groundball friendly starter in major league baseball.
On to the evidence.
In order to look at a game by game breakdown, I’m using ESPN’s numbers from their game logs.
Roy Halladay:
Season G/F rate: 2.59
Single Performance Low: 1.5
Single Performance High: 7.5
Median Performance: 2.60
Standard Deviation: 1.52
Derek Lowe:
Season G/F rate: 2.99
Single Performance Low: 1.13
Single Performance High: 12
Median Performance: 3.12
Standard Deviation: 2.49
Brandon Webb:
Season G/F rate: 4.25
Single Performance Low: 1.33
Single Performance High: 16
Median Performance: 4
Standard Deviation: 3.48
John Patterson:
Season G/F rate: 0.55
Single Performance Low: 0.08
Single Performance High: 1.4
Median Performance: 0.55
Standard Deviation: 0.34
Scott Elarton:
Season G/F rate: 0.67
Single Performance Low: 0.15
Single Performance High: 2.25
Median Performance: 0.75
Standard Deviation: 0.60
Pedro Martinez:
Season G/F rate: 0.82
Single Performance Low: 0.15
Single Performance High: 4.33
Median Performance: 0.75
Standard Deviation: 0.87
That gives us six pitchers, three of whom are severe groundball pitchers and three of whom are severe flyball pitchers. Among the interesting things I’d observe about the data (and still being aware that this is all dealing with small sample sizes).
There is far, far more deviation among the groundballers than among the flyballers. I actually expected that to be reversed. My theory, going into this, was that an extreme groundball pitcher would almost always be an extreme groundball pitcher. While it is true that Halladay, Lowe, and Webb have not had one game all season where they allowed more fly balls than ground balls, their levels of groundballness was all across the board. I didn’t expect to see that.
The median correlates very strongly with the average, which is true in almost any statistical study you’re going to do. However, I was surprised at how closely they matched.
Pedro Martinez was, until last year, a groundball pitcher. He’s now one of the more extreme flyball pitchers in the game. I have no idea why.
So, what does this all have to do with Felix?
Well, the 12-0 mark he put up in Detroit, I believe, is a huge clue that Felix is going to be an extreme groundball pitcher. You don’t put up something like that on accident, small sample size or not. You’ll never see John Patterson go 5 innings without allowing a flyball. It just isn’t going to happen. But his ratio on Tuesday night of 12-6 is much more human, something that would fit in at the low end of the spectrum for Halladay, Lowe, and Webb. Also, from my notes, most of those fly balls came later in the game, when it could be assumed that Felix was tiring slightly. Perhaps, if he had gone more than 5 innings in Detroit, his ridiculous 12-0 ratio would have come out as something more like 15-5.
Obviously, Felix will sort this all out by himself when he takes the mound the rest of the year. However, I believe, based on his two starts, we can with some assurances extrapolate that, going forward, Felix Hernandez is going to be one of the most extreme groundball pitchers in major league baseball. Whether he’s Roy Halladay or Brandon Webb, we can’t know yet. But we should be certain that he’s not a flyball pitcher.
So, why does this matter? Well, here’s the thing about groundball pitchers who also strike out a high percentage of hitters; they are, with almost no exceptions, all-stars. The combination of groundballs and strikeouts is death to offenses.
People will talk about Felix needing to improve his command of his fastball and gushing about the potential of a 19-year-old. I say, right now, Felix Hernandez, with the heavy sink of his 2-seam and the nasty break on his curve, is a lights out major league starting pitcher, no improvement necessary. The potential is outstanding, but he’s arrived. He is a frontline major league starting pitcher right now.
Comments
65 Responses to “Felix Hernandez, Groundball Machine”
37 – Players can argue to ignore their service time if they make some sort of special contribution – like if Felix won a Cy Young Award. Then he’d just get compared to other Cy Young calibre pitchers, and we’d end up paying him $12 million.
Brian,
No, the ESPN numbers (which are the ones I used) are groundball and flyball, as determined by Stats Inc. They are different from groundout/flyout, and should include all balls hit on the ground or in the air, regardless of outcome.
37 – Players can argue to ignore their service time if they make some sort of special contribution – like if Felix won a Cy Young Award. Then he’d just get compared to other Cy Young calibre pitchers, and we’d end up paying him $12 million.
Sort of. You get to argue your case on your merits (“including but not limited to overall performance, special qualities of his career contribution…”) but the arbitrator’s instructed to pay “particular attention, for comparative salary purposes” to other players with equal service time.
The player can say that that’s irrelevant because of “special accomplishment” which merits comparing to other players regardless of service time or contract status, and the arbitrator then considers this as part of any other argument. Even if you win a Cy Young award, and wish to introduce the salaries of other Cy Young winners, that argument doesn’t, once won, mean that service time is no longer relevant at all. It only means that you evidence on equivalent salaries will be weighted with the service time arguments in determining which of the two numbers is closer.
#44… Jim, I don’t think we really disagree here. The loyalties of a player to a franchise and vice versa is flimsy, at best. It’s not necessarily a good idea for either side to bank on it. I think building a good franchise around a player as opposed to sinking all future contract money into that player (the Mike Sweeney approach) is a better way to go. In the ARod case, his decision to abandon the M’s in 2000, still left the team in pretty good shape in 2001, wouldn’t you say? It’s like hedging your bet.
I’m simply going to enjoy these next 6 years and not worry too much about what happens after that. And I will try to hold my tongue and avoid the Monday morning quarterback recriminations against management if we actually lose him on the 7th year.
Thanks, Dave, I guess I should have reviewed for your source citation, and looked up their definitions. Sorry for the inconvenience.
I sure hope that Scott Boras doesn’t get Felix’s phone number from Adrian.
Actually, Bavasi seems to have a pretty good working relationship with Boras. It is one thing that I am definitely greatful for in regards to Bavasi. If Boras continues to attract top talent to his agency, then having a GM that knows how to deal with that agency is definitely a plus.
Dave – Simple suggestion on the s.d. problem. Divide mean by s.d. and use those to judge the variability. Samples with higher means many times, and specifically in this instance, will naturally have higher s.d.’s because larger numbers are involved. Alternatively, you could calculate F/G ratios and it would look like the flyball pitchers were more variable.
On nicknames: The fastball has one already–“ultima ratio regum.” And that sticks even if the curve becomes his out pitch.
Or, for short, “The Argument”. Like the Fugazi song (and album)
Actually, I also envision “Ultima Ratio Regum” as two-strike chant, sung with clearly Latin pronunciation, and the proper Gregorian descending pitch interval on the final syllable of each word:
“Ool-tee-mah . . . rah-tee-oh . . . ray-goom”
Oh yeah.
“Small sample size police!” – wow! A term whose time has come. And on the heels of “King Felix.”
David, you’re on a roll.
Well, it someone hasn’t mentioned this already, this may put more pressure on Bavasi to keep a very good defense infield. I think there is no way he’ll get a power guy with shitty defense for the infield. He’s gotta keep his stars happy. So I see this having a trickle down affect on the roster. Morse surely is gone from SS and Willie will never become the 2B. Morse may become the 2B but I bet the FO will hope to find something lucky: a 2001-Boone version. Any “big” bats will probably be at the expensive of Reed’s or Doyle’s (unfortunately) playing time. This in turn could mean Raul back in left if Bucky proves to be a cheap power bat DH they are looking for. The 2006 roster will be very interesting.
Thanks for the thread, Dave; this is the only place in ‘town’ one can follow a discussion of this kind.
Looking at #35, Dave, it would appear that lefties actually hit more flyballs than groundballs off Felix El Rey. Which makes sense, since lefthand batters tend to hit the low ball better, not least because the also get a better look at the pitch. OTH, lefties also struck out more. The sample isn’t large to be sure, but this suggests something interesting to come: Felix has a slider, and that pitch may very well just eat. lefties. up. He annihilates righties now. With a slider boring in on the lefthand batters . . . This cat’s got it all.
“He’s a frontline major league starting pitcher right now.”
Oh yummy! I’m going on Monday.