Felix Hernandez and Missing Velocity

When Felix Hernandez broke into the Majors as a 19-year-old, he threw the ball really hard. Here’s the Pitch F/x plot of his start against the Oakland A’s on April 2nd, 2007, the first for which we have Pitch F/x data available.

The fastball was 95-100, the slider was 89-92, the change-up was 86-89 (ignore the “FS” labels, as the algorithm wasn’t so good back then), and even his “slow curve” was 82-87. Felix was the embodiment of a power pitcher, and while his three off-speed pitchers were all notable in their own right, Felix’s mid-90s two-seamer was his defining pitch. There just weren’t many guys in the game that could run a sinking fastball up there at 95 MPH, and that pitch helped him run a 60.8% GB% that year.

Now, here’s Felix’s Pitch F/x plot from his start against Oakland on Saturday night.

The change-up is still in the upper-80s, while the slider and curve are both down a couple of ticks on average, but not too terribly far from where they were five years ago. However, the fastball is 90-93, and because of its close proximity to the change-up, it’s basically indistinguishable in the chart. Instead of two distinct clumps, there’s just now one big mess of pitches in the 90 MPH range with similar movement and velocity. In fact, the similarity of the two pitches basically broke the the Pitch F/x algorithm, as it ended up classifying 47 of his 102 pitches as change-ups due to an inability to tell the change from his fastball.

In reality, if you just look at pitches between 86-90 MPH and break out the obvious sliders, it looks like he threw 26 change-ups on Saturday. Despite general acceptance of the belief that a change-up needs a lot of velocity separation from the fastball in order to deceive hitters, Felix’s change on Saturday night was still as devastating as ever. Of the 26 real change-ups, he got swinging strikes on eight of them, a staggering 31% whiff rate. On the 76 fastball/slider/curves he threw, he got three swinging strikes combined.

It’s not just the lack of swinging strikes that set his fastball apart on Saturday night, however. As noted, Felix came into the league as an extreme groundball pitcher. As his velocity has declined slightly over the last four years, he’s settled in as more of a 50-55% groundball guy, but has still been able to get a lot of grounders with his sinker.

On Saturday, though, Felix allowed 19 balls in play (excluding Eric Sogard‘s bunt), and only five of them hit the turf. The 26.3% GB% wasn’t a career low in a start for Hernandez, but he’s only gone through a start with a GB% below 30% on five other occasions in 6 1/2 years as a big league pitcher. It is an unusual outcome for Hernandez, to say the least.

On it’s own, though, it doesn’t really mean anything. Felix had a start against San Diego last year where he induced only four ground balls on 15 balls in play, and yet it was one the best performances of his career. While GB% stabilizes very quickly, one game variance almost certainly isn’t predictive of anything. And, we also have to keep his opponent in mind – the 2011 Athletics had the second lowest GB rate of any team in baseball, and then added extreme fly ball hitter Josh Reddick and looking-like-a-flyball-guy Yoenis Cespedes to their roster over the winter. The opposing batter certainly has an effect on batted ball trajectory, and the A’s aren’t a team that hit the ball on the ground a lot, so we have to reset our expectations for what a normal GB% for Felix against them might look like.

When combined with a significant reduction in velocity, though, it’s something to keep an eye on. A 95 MPH sinker is just more likely to get a ground ball than a 90 MPH sinker, and if Hernandez’s fastball is going to be more low-90s than mid-90s, we would expect him to get fewer ground balls as a result. However, the if part of that statement is still a big if, as early season velocity is generally lower than mid-season velocity, and Hernandez might very well come out throwing 94 again in his next start. Back in 2008, Justin Verlander started the season throwing 92 and ended it throwing 95. Madison Bumgarner famously lost his fastball for a few months at the end of 2009 and beginning of 2010, then had it reappear later in the season. Declines in April velocity are not always harbingers of significant problems to come.

However, Felix wouldn’t be the first power pitcher to go through this kind of dramatic transformation. Tim Lincecum came into the Majors with a fastball that averaged 94 and regularly got into the high-90s, but has morphed into a pitcher who relies heavily on a change-up that is among the game’s best as his fastball has moved more into the low-90s over the last couple of years. Hernandez’s secondary stuff is good enough that he doesn’t have to rely on a big time fastball in order to get hitters out, and like Lincecum, he could probably adapt his game to play more off the soft stuff if his fastball remains less of a weapon that it has been previously.

While you can throw most early season performances out the window as completely non-predictive, significant changes in velocity are one thing that are at least worth monitoring going forward. If his fastball comes back and he gets the A’s to hit the ball on the ground regularly on Friday night, then this may all be relegated to the dustbin of weird April flukes. But, considering that Felix spent all of spring training in the 88-92 range and hasn’t flipped a magic velocity switch with the start of the regular season, this will continue to be a story until he starts sitting at 94 again.

If the velocity never does come back, we’re going to have one very interesting experiment in change-up effectiveness, especially if that pitch continues to be the primary weapon he uses when going for a punch-out. Even Lincecum’s change has had 7-8 MPH of separation from his average fastball speed. Can Felix keep racking up strikeouts with a change-up that is so close in velocity to his fastball? Will he even need to?

We don’t have the answers to these questions, but it is a situation worth keeping an eye on.



Print This Post



Dave is the Managing Editor of FanGraphs.


Sort by:   newest | oldest | most voted
Trev
Guest
Trev
4 years 4 months ago

Justin Verlander’s 2008 was agonizing for Tiger fans. Nothing went right that year. I hope King Felix doesn’t go through a season like that. His just too fun to watch.

Cozar
Member
Cozar
4 years 4 months ago

I don’ have any hard evidence to back it up, but it frequently seems like pitchers pitch poorly when they have large leads. It’s almost like they don’t feel the need to push their stuff to the max, then when they start getting hit, they’ve already lost the mental game and can’t turn it around.

It might be that the batters just started getting a read on Felix, but I can’t help but think it wasn’t purely coincidental that Felix started to fall apart after he got a 7-0 lead. I’d love to see a Fangraphs article to support or refute the “big lead” effect.

Haha no
Guest
Haha no
4 years 4 months ago

Jack Morris has been spiking your drinks.

Cozar
Member
Cozar
4 years 4 months ago

Maybe, but in my defense, Felix seems to pitch worse when his team scores 6 or more runs:

0-2 runs – 2.94 SO/BB, ..656 OPSA, 3.03 ERA
3-5 runs – 3.15 SO/BB, .630 OPSA, 2.85 ERA
6+ runs – 2.90 SO/BB, .724 OPSA, 4.20 ERA

And he’s a lot worse when there is greater than a 4 run differential:
Margin > 4 runs: 2.37 SO/BB, .853 OPSA,
1-3 runs (estimated average): 2.90 SO/BB, .650 OPSA
Tie: 2.73 SO/BB, .654 OPSA

This latter comparison is less useful, as there is obviously a causation issue (pitching terrible would create a 5+ run margin) as well as a low small sample size (only 42 games). But, I think there is enough there it would be interesting to see if pitchers drop velocity or lose a little command when they have big leads.

Cozar
Member
Cozar
4 years 4 months ago

Interestingly, Jack Morris doesn’t display at all the kind of disparity that Felix does.

Jack’s career SO/BB ranges from 1.66 (tie) to 1.99 (>4 run), but his OPSA ranges from .684 (1 run) to .699 (>4 run) a much lower disparity.

His run support numbers are very interesting. They have the same trend for OPSA and ERA (best at 3-5 runs) but he has a higher SO/BB with 6+ runs and seems to pitch as well with 6+ runs as he does with 0-2.

0-2 runs – 1.74 SO/BB, .710 OPSA, 3.98 ERA
3-5 runs – 1.70 SO/BB, .667 OPSA, 3.54 ERA
6+ runs – 1.99 SO/BB, .711 OPSA, 4.24 ERA

Maybe what we are seeing here is that some players do perform better in high-pressure situations while some pitch the same regardless of the amount of pressure. That might suggest some players have a closer mentality and some don’t need it.

CircleChange11
Guest
CircleChange11
4 years 4 months ago

Okay, so Felix didn;t pitch that well the 2 times the mariners scored 6 runs or better. *grin*

Actually, if the mariners scored 6+ runs perhaps there was something in the environment (wind?) that was affecting the overall offensive production on the day.

If the M’s are scoring a bunch of runs and Felix is allowing a bunch of runs, my initial feeling would be there’s something going on with the environment … and not global warming.

Cozar
Member
Cozar
4 years 4 months ago

Not exactly, Circle. It’s a rational hypothesis that pitchers might not try as hard when the they have a large lead (or a huge deficit) in order to save their best stuff for another day when it matters more.

I’m wondering if 1) there is any truth to that and 2) if that is what happened to Felix.

Notably, he was pitching a 1 hitter until he had a 7-0 lead. With a 7-0 lead he wouldn’t be quite as worried about giving up a few hits or runs. That might have changed his mental approach which resulted in giving up more hits.

davisnc
Member
Member
davisnc
4 years 4 months ago

It’s strange to me that the diminished velocity hasn’t just taken the change-up down correspondingly. Obviously it’s a very effective pitch historically, which leads me to believe that his arm action is probably very similar to his fastball arm action, since that’s a basic requirement of an effective change-up. I’ve watched a decent number of his starts, and this certainly seems to be true observationally, but then again I’m not a Major League hitter or pitching coach. If his mechanics are allowing him to generate that kind of velocity on the change, (89.2 so far this year vs. 89.4 last year), it seems like everything should still be there to pump the fastball velocity up.

I realize I’m not adding much here. Just a bit puzzling.

vivalajeter
Guest
vivalajeter
4 years 4 months ago

Maybe the arm strength is there but he has a different grip, which alters movement and velocity?

davisnc
Member
Member
davisnc
4 years 4 months ago

Yeah, if it came out that he was experimenting with a new grip or something that would absolutely explain it. I’d guess that if that were the case, we’d have heard about it by now, since he’s probably sick of answering questions about his velocity.

I know he lost a fair amount of weight over the offseason, but if that were a significant factor, I’d expect the CH velocity to decline as well, which it hasn’t. Obviously I’m grossly oversimplifying the mechanical side of things (e.g. maybe the velocity on those 2 pitches comes from different places, if he doesn’t push off as hard for the change or something). Meh. My guess is he’ll be fine.

Mr Punch
Guest
Mr Punch
4 years 4 months ago

It appears that although there’s less separation between fastball and change in terms of speed, there’s more separation in terms of movement (and not necessarily in the way I might have expected).

Corey
Guest
Corey
4 years 4 months ago

The thing that makes me want to dismiss this long term (other than just the raw hopes of a Mariners fan) is that he looked like Felix Hernandez in Japan. I didn’t see it, but I heard he was lacking velocity in his last spring outing as well. This is new not just compared to last season, but also compared to 2, 3 weeks ago when Felix Hernandez looked like Felix Hernandez. The other thing that I wonder about is the very focus of this article. The change-up velocity. I too noticed the other day that Felix’ changeup was indistinguishable from his fastball, to the point that I somewhat question the ability to identify “real” changeups, but let’s take you at your word that there were 26 identifiable “real” changeups. Doesn’t that indicate that Felix hasn’t actually lost arm speed? A good changeup is supposed to maintain arm speed while achieving a slower speed due to the way the ball exits the hand. If Felix’ changeup is still coming in at the high 80s-90mph range doesn’t that indicate that he’s still able to generate the same arm speed as before? So maybe he’s just gripping the ball too tightly with his fastball or for some reason letting it fall into the back of his hand (like a changeup)? I thought it was a disturbing start due to the reduced fastball velocity, and I thought it was really stupid of Wedge to let him hit triple digits in pitch count for that reason, but I’m not convinced this indicates anything long term wrong with Felix. If I had to bet on it I would bet either a) jet-lag coming back from Japan or b) something subtle mechanically (like grip).

night_manimal
Guest
night_manimal
4 years 4 months ago

Pitch classification seems off by quite a bit. Did the cutter and change up’s get mixed up? Who throws a change the same speed as a fastball?

Corey
Guest
Corey
4 years 4 months ago

As a fan watching the game, it wasn’t obvious that he threw ANY changeups, so either the classifications are right, or he didn’t throw any at all.

JW
Guest
JW
4 years 4 months ago

I may be wrong, but it seems like Felix’s pitch speed may already be improving. Its hard to say for sure because of the lack of Pitch F/x data from the start in Japan, but the Reports coming out then, and based watching Felix’s last spring start against the Rockies, it seems like he was sitting in the 89-90 mph range on his fast ball and occasionally touching 91, while in his game against the A’s he was sitting mostly above 90 and touched 93 a couple times.

Mark
Guest
4 years 4 months ago

The concern I have with Lincecum is that he seems to have lost even more velocity this spring and during his first starter and he’s scrapped the slider which was a good pitch for him last year to save his elbow. Those are ominous signs.

Young Whipper-Snapper
Guest
Young Whipper-Snapper
4 years 4 months ago

I know Timmy came out and said that he isn’t going to throw his slider as much as he did last year, but in his first start this season he threw quite a few sliders, which surprised me. I don’t have the exact number in front of me, but I wouldn’t be surprised if the number was right in line with the 17% mark he put up last year.

McE
Guest
McE
4 years 4 months ago

Maybe it’s just rationalization, but I get the impression that Felix has been saving himself in Spring Training and the radar gun will start to reflect that going forward. Michael Pineda was also known to “conserve” velocity in his time with Seattle, and I assume he learned that trick from someone.

I’m crossing my fingers that before long Felix will look like the King again, and I can go back to wishing on a 2013 rotation of Felix, Walker, Hultzen, Paxton, and a fifth guy too.

Phantom Stranger
Guest
Phantom Stranger
4 years 4 months ago

I think this is closer to the truth than diminished stuff. Most of the game’s top starters realize they are going to throw 220+ innings in a season, not counting playoffs. Even guys like Halladay and Cliff Lee don’t throw their max effort stuff this early in the season. It is what separates them from pitchers like Josh Johnson, who only know how to throw as hard as they can in every start.

smb
Guest
smb
4 years 4 months ago

Josh Johnson can’t seem to get through a season, and Felix, as we know, is made of cast iron, but even cast iron can break if you heat it up or cool it down too fast in succession, right? If that’s why he takes his time warming up into the season, then by all means…

ben
Guest
ben
4 years 4 months ago

“Even guys like Halladay and Cliff Lee don’t throw their max eff”ort stuff this early in the season.”

Can I ask what this statement is based on?

CircleChange11
Guest
CircleChange11
4 years 4 months ago

Yeah, Halladay is notorious for April complete games … especially in Toronto.

Ian
Guest
Ian
4 years 4 months ago

As a huge Felix fan, it is disturbing to see the velo drop, but at the same time, this seems like a common issue with pitchers as they progress in their careers. Clayton Kershaw is another example of a pitcher who was reported to sit 95-96 coming up and he’s much more in the 91-94 range these days. And obviously he had a tremendous season last year. Verlander is so unusual since his velocity actually increases as the game goes on, that shows some of these guys can actually conserve the arm speed if they want to.

The only other thing I would say about Felix is that he seems to rely too little on his fastball. While his offspeed stuff, especially the changeup, can be devastating, I’d love to see him use the fastball more.

McE
Guest
McE
4 years 4 months ago

It took M’s fans several years and one open letter from none-other-than Dave Cameron himself to get Felix to use his fastball less. Let’s not go putting ideas in his head, now, hmmm?

http://blog.seattletimes.nwsource.com/mariners/2007/07/hernandez_keys_40_win.html

Brandon Warne
Member
Member
4 years 4 months ago

Is it possibly more of a self preservation thing?

Tim F
Guest
Tim F
4 years 4 months ago

Is this even remotely true?

“A 95 MPH sinker is just more likely to get a ground ball than a 90 MPH sinker”

I’ve heard that sinkerballers pitch better as they tire. No clue if either is true, but not going to take the above at face value.

Tyler
Guest
Tyler
4 years 4 months ago

It is. Sinker-ball pitchers do not want to be “too strong” during their starts as they’ll leave sinkers up if they are. Andy Pettite used to frequently say after a bad start that he was too strong and his off-speed stuff was left up as a result.

Gregory H
Guest
Gregory H
4 years 4 months ago

Does Pitch f/x, like most radar guns, measure pitch velocity shortly after the release? Or does it compute velocity both shortly after the release and as the pitch crosses home plate, and then compute an average velocity? In either scenario, change-up velocity as recorded is deceiving.

Felix Hernandez could still have a very effective change up if his change up loses speed far more rapidly than his fastball, and any change-up should. In fact, having a similar velocity out of the hand would arguably make his change more effective because it would be practically indestinguishable to a hitter from the fastball. Felix throws his change with a standard two-seam circle change grip. The beauty of the circle change is that it should have a very similar velocity shortly after the release point when compared to the pitcher’s fastball.

Lincecum throws what he calls a modified splitter grip for his change up. I’ve seen photos of this grip, and it really is more of a forkball grip. Unlike Bruce Sutter, Lincecum rests his thumb on the front seam instead of the back seam, and Lincecum’s grip is wider than a classic splitter grip. Since Lincecum uses such a wide grip, I would expect a greater difference in velocity shortly after release when compared to his fastball.

Phrozen
Guest
Phrozen
4 years 4 months ago

The Pitch f/x system tracks velocity across the entire pitch; giving both release and platye velocities.

Have no idea which if either is presented here.

Nathaniel Dawson
Guest
Nathaniel Dawson
4 years 4 months ago

Where does this come from? Why would a change-up lose more velocity on it’s trip to the plate than a fastball?

It doesn’t seem logical — if anything, a fastball thrown at a higher velocity than a change should lose more from the pitcher’s hand to the plate because air resistance is higher with increased velocity.

monkey business
Member
monkey business
4 years 4 months ago

So what is causing the differential in deceleration?

SAS
Guest
SAS
4 years 4 months ago

As the pitch f/x velocity data piles up over years, it will be interesting to see what we learn about a pitcher’s velocity of a career. It seems like a lot of pitchers come up throwing in the mid to high 90’s, but few seem to sustain that.

Luke M.
Guest
4 years 4 months ago

Add Tommy Hanson to the list. He’s coming back from injury right now so he’s working 87-91, but even at his best he tops out around 92-93 now.

Shaun Catron
Guest
Shaun Catron
4 years 4 months ago

Cozar saying that Felix might have struggled because the Mariners scored 6+ runs. Felix lost veloctiy because the Mariners hit some homers and gave him run support? what?

Cozar
Member
Cozar
4 years 4 months ago

No, I’m not suggesting he lost velocity because he got more run support, I’m suggesting it’s possible, because he had a huge lead, that he simply wasn’t trying as hard.

wpDiscuz