Why should pitchers resolutely throw fastballs 60% of the time?

[rotundlio]

They're the easiest pitches to hit by far. I subscribe to the belief that it's the pitcher's job to make hitting as difficult as possible, and you wouldn't seem to accomplish that by repeatedly dialing up the straight stuff. Fastballs appear on the surface to be ineffectual.
 
Take Yu Darvish for example. He throws eight! pitches per BrooksBaseball; for the purposes of this exercise I'm going to discount sinkers and cutters, focusing instead on the fourseam versus the offspeed. Brooks gives me this.
 

 
FanGraphs paints a similar albeit disjointed picture.
 

 
Pitch 1 is the fastball. Why should Darvish throw this pitch far more extensively than any other? What compels him to do that? If we're analyzing his repertoire based solely on this data, we're led to believe that Yu is better served abandoning Pitch 1 entirely.
 
It oughtn't be a question of command. In many cases pitchers pound the zone more frequently with their breaking stuff, especially when you consider that it's mostly thrown while ahead in the count. If you place a tin can atop your head and stand sixty feet-six inches from 2013 Matt Harvey, you'd be in little danger. Strike throwing isn't the crux of it. Hitters are much more likely to chase breaking pitches outside the zone in any event.
 
To my knowledge, there isn't any assertive evidence for benders being harder on the arm. FanGraphs posits that pitching itself is one long injury. They figure curveballs to be harmful, but note the correlation between fastball usage/velocity and UCL tearing. I'm receptive to the idea that fastballs are easier on the body, but I have no proof. At any rate, football players pound their brains to mush in pursuit of their Ws.
 
Obviously, there's an argument to be made that fastballs help "set up" other pitches, particularly changeups due to arm action, etc. I don't dispute it, but Darvish's secondary offerings would have to worsen pretty drastically to be rendered less effective than the fourseam. Jake Arrieta is the only pitcher I can recall who throws a breaking pitch as his #1. His heater sits at 94-95 for good measure, and still it pales in comparison. Moreover, shouldn't this phenomenon work in reverse? If breaking pitches play up to the fastball, why isn't Darvish's of all fastballs likewise made a better pitch?
 

 
Darvish, Felix, Kershaw, Lester, Sale: the first five pitchers that came to mind, and generally poor examples.
 
Might this be the future of our game? In 2250, when baseball is the last remaining professional sport, might people look back and laugh at the days when most pitches didn't even bend? I mean... look how easy it is to hit. And what of the next crop of Y2K Anthony Davis pitching prospects? Are we approaching the era of 98 mile-per-hour sliders? I'm reminded of this video.
 
I'd love your feedback.
 
 
P.S.  I've always wanted Sons of Sam Horn to realize how badly they've needed Out of the Park Baseball in their lives.

 

[CarolinaBeerGuy]

It seems to me that the high percentage of fastballs is what makes the offspeed stuff so effective. If the batters knew they were going to get a high percentage of offspeed pitches, their approaches would change and they'd start pounding those pitches.

 

[Lose Remerswaal]

Throwing too much offspeed stuff causes nasty torque on the arm (see: Tanaka, Masahiro)?

 

[joe dokes]

It oughtn't be a question of command. In many cases pitchers pound the zone more frequently with their breaking stuff, especially when you consider that it's mostly thrown while ahead in the count.
 

 
I think you are overestimating the collective ability of pitchers to throw strikes with breaking stuff, as opposed to fastballs.  What makes great pitchers great is the ability to throw breaking stuff for strikes when behind in the count.  And when ahead in the count, good pitfcers often throw breaking pitches that are *not* strikes, getting hitters to chase.

 

[Fred not Lynn]

And don't forget not all fastballs are equal. A well located fastball, even without ridiculous velocity, can be a very effective pitch. It's not like when a pitcher throws a fastball for a strike it's over the heart of the plate most of the time. Plus, you need to throw enough fastballs so your off-speed pitches have a speed to be off of.

 

[joe dokes]

And don't forget not all fastballs are equal. A well located fastball, even without ridiculous velocity, can be a very effective pitch. It's not like when a pitcher throws a fastball for a strike it's over the heart of the plate most of the time. Plus, you need to throw enough fastballs so your off-speed pitches have a speed to be off of.

 
As Warren Spahn is credited with saying: "Hitting is timing.  Pitching is upsetting timing"

 

[Fred not Lynn]

I guess the other way to put it is if all you ever throw are off-speed pitches, they're not off-speed, they're just slow.

 

[Comfortably Lomb]

Throwing too much offspeed stuff causes nasty torque on the arm (see: Tanaka, Masahiro)?

 
Isn't the buzz around him that the splitter specifically is a problem when thrown too much? I've heard some people say it's the slider too. Others the curve but only for younger arms. There are those out there that thing fatigue influencing mechanics is the real culprit. I'm not sure anyone knows for sure. Baseball players are a superstitious lot and highly susceptible to suggestion.

 

[rotundlio]

Throwing too much offspeed stuff causes nasty torque on the arm (see: Tanaka, Masahiro)?

 
Are we sure? Schilling relied on splitter-slider, and that guy made 436 starts. I think Tanaka's workload in high school is just as readily to blame. It's faintly disturbing in Japan—teenaged Daisuke threw 148 pitches one night and 250 the next. Seventeen-year-old Masahiro threw 742 across two weeks after ostensibly pitching the team's way into the tournament. And what's with all the young, electric flamethrowers requiring Tommy John?
 

What makes great pitchers great is the ability to throw breaking stuff for strikes when behind in the count. 

 
...Is this exceedingly uncommon? Does that truly denote greatness? I thought this sort of thing made Major League pitchers Major Leaguers. This is the most infinitesimal topmost sliver of sphere throwers in the world, but I could be wrong. I can't find pitch data by count.
 

And when ahead in the count, good pitfcers often throw breaking pitches that are *not* strikes, getting hitters to chase.

 
Even despite this, breaking balls catch a lot of zone. League average PITCHf/x Zone% on FanGraphs is 45%. Darvish's slider sits at 46%; Felix's slider and curve 47% and 49%; Sale's slider 46%. Just throwing darts... Archer's slider, 46%; Buchholz's curve/slider, 48%/46%; Latos's curve, 46%. These figures would leaven if they were primary offerings. In every case hitters make worse Z-contact against these than they do against fastballs. Breaking balls are chased outside the zone 50-100% more often regardless, with hugely worse O-contact. I don't blindly accept that these pitches are best used sporadically as complementary pieces.
 

It seems to me that the high percentage of fastballs is what makes the offspeed stuff so effective. If the batters knew they were going to get a high percentage of offspeed pitches, their approaches would change and they'd start pounding those pitches.

 
Possible. Arrieta's the one example I know; he threw more sliders than fastballs last season. Even sitting 95, the fastball yielded laughably worse results. Koji throws 89 with 40% splitters, but that remains completely unhittable. (Actually a case of the fastball playing up.) In any case, I wouldn't advise Jose Quintana to quit throwing his fastball. I'd suggest that, if fastballs are unquestionably the least deceptive pitches and the easiest to hit with a bat, curve-splitter-slider-screwball could prove a better archetype. If you touch 95 with one of those, as future generations might, there'd be no question!
 
Z-contact on breaking pitches was higher than I'd anticipated—nearly comparable to that of fastballs, really, on the whole. But Z-contact in general is extremely high... 87%.
 
Aroldis Chapman's fastball is impossible to see, and still the slider profiles better. What's that portend? You can still change speeds without throwing ramrod straight. Velocity's climbed to 92 on average: slightly faster than Harvey's slider (65% contact).

 

[Clears Cleaver]

I'd like Schilling's view on this. I think he's made several points about how the fastball is the most important pitch.
 
What may skew your numbers is when fastballs are throw in the count? versus when off-speed pitches are thrown? Although the historical "get ahead with fastball and out with breaking pitches" mantra might be what the real argument of this is when all is said and done. But generally, fastballs thrown when behind in the count (As they are often done) are likely more hittable than offspeed and breaking pitches thrown when ahead in the count (as they often are).

 

[rotundlio]

That definitely colors the stats, but how prolific can it be? Kershaw's slider is three times as likely to be missed and twice as likely to be chased. At some point it's just a more difficult pitch to hit.

 

[Rovin Romine]

And don't forget not all fastballs are equal. A well located fastball, even without ridiculous velocity, can be a very effective pitch. It's not like when a pitcher throws a fastball for a strike it's over the heart of the plate most of the time. Plus, you need to throw enough fastballs so your off-speed pitches have a speed to be off of.

 
Shouldn't we really be looking at the FB and the change up as a "single" pitch?  Provided you've got a pitcher that throws both.  While not thrown as often as the FB, it looks like the change is a pretty effective weapon for a lot of the pitchers on the list in the first post.  

 

[Idabomb333]

 
I think you are overestimating the collective ability of pitchers to throw strikes with breaking stuff, as opposed to fastballs.  What makes great pitchers great is the ability to throw breaking stuff for strikes when behind in the count.  And when ahead in the count, good pitfcers often throw breaking pitches that are *not* strikes, getting hitters to chase.

This should be testable, right?  Has anyone ever run a comparison between, say, % of fastballs per total pitches thrown vs. % of breaking balls thrown for strikes by pitcher?  If pitchers throw fastballs because they have to do so to throw strikes, then those who can control their breaking balls better would throw them more.

 

[Al Zarilla]

Some good reasons already but also pitchers can generally pump fastballs over the plate more easily than their other pitches. The term "get it in fastball" comes to mind, and the data should say, I would think, that pitchers throw more fastballs than any other pitch when they're in danger of walking a batter. Should be the same when the count is 0 - 0, i.e., more first pitch fastballs.
 
Edit, I said somewhat the same thing as Idabomb.

 

[EricFeczko]

The way you are examining the fastball here is incomplete. The simplest explanation is that the fastball is put in play the least. Here's Darvish Yu for every pitch thrown:

Yu Darvish Pitch Outcomes - from 03/30/2007 to 03/10/2015:

Pitch Type	Count	Ball	Strike	Swing	Foul	Whiffs	BIP	GB	LD	FB	PU	HR
Slow Curve	494	43.52%	37.65%	29.55%	7.89%	11.34%	10.53%	3.85%	2.83%	3.24%	0.61%	0.61%
Fourseam	3087	40.01%	30.45%	37.51%	16.33%	8.65%	12.93%	3.98%	3.14%	4.24%	1.55%	0.68%
Curve	621	36.72%	34.30%	50.40%	14.81%	21.74%	13.69%	7.25%	3.06%	3.06%	0.32%	0.48%
Slider	2021	29.89%	39.98%	49.73%	15.04%	20.04%	14.75%	5.54%	3.32%	3.81%	2.08%	0.45%
Split	341	42.82%	26.39%	50.73%	14.66%	19.94%	16.13%	9.09%	4.40%	2.35%	0.29%	0.00%
Cutter	1546	31.69%	29.82%	50.19%	21.15%	12.68%	16.82%	8.80%	2.98%	3.95%	1.10%	0.58%
Change	211	47.87%	22.27%	41.23%	12.32%	11.85%	17.06%	10.43%	4.74%	1.42%	0.47%	0.95%
Sinker	1231	39.97%	22.01%	44.27%	17.63%	7.31%	19.50%	9.91%	4.39%	4.22%	0.97%	0.41%

A more complicated explanation is that it is much harder to hit a faster pitch than a slower pitch. Essentially, you have less time to react to a fastball than any other pitch. A 95 MPH fastball reaches home plate in ~434 milliseconds. An 85 MPH slider reaches home plate in 485, while an 80 MPH curve reaches home plate in over 515 milliseconds. However, what also matters is the amount of time the player has before he's required to decide to swing the bat. It takes approximately 50 milliseconds after the decision to start the swing (feet and arm motion need to be synced, and it takes ~50 milliseconds for a neural signal to reach the feet from the CNS). The bat needs to travel about a meter before connecting with the ball; depending on bat speed, this can add another 25-50 millseconds. So the player needs to decide to initiate a swing when the ball is ~100 milliseconds from where he thinks the ball will be.
 
This means two things: 1) the margin of error for determining the future location of a fastball is smaller, due to the increased velocity. This leads to more foul balls. 2) A player only has ~300-350 milliseconds to discriminate whether the pitch is a fastball before deciding to swing (actually they have 300-350 seconds to discriminate between two-seamers, four-seamers, and cutters). They have an additional 50-100 milliseconds to discriminate between offspeed pitches. Because fastballs are thrown more frequently, this biases the hitter towards using less time to make a decision.
 
It will also bias the hitter to hit fastballs better. The reverse would be true if you flipped the pitches (i.e. pitch more offspeed than hard pitches), only in the reverse situation you are generally giving the hitter more time to make a decision.

EDIT: Meant to say "a" more complicated explanation. Obviously, other factors with respect to the pitcher (e.g. easier to control a fourseam fastball) are valid explanations.

 

[rotundlio]

That's fascinating stuff. Thanks.
 
Still, Felix's fastball is put into play relatively often. So are Kershaw's, Sale's, and Strasburg's. While these guys are certainly exceptions to the rule, Yu Darvish is perhaps the most extreme example of this phenomenon in all of baseball. It's worth noting also that hitters are far more likely to offer at sliders, splitters, and changeups. On a per-swing basis, I'm not positive fastballs really are put into play substantially less often.
 
Furthermore, breaking pitches induce much weaker contact. Their groundball rates are markedly higher, their line drive rates appreciably lower. It's a limited sample I've got here, but fastballs would appear to be relatively easier to hit for home runs. I can't find leaguewide ratios, but among the previous sample of Felix, Sale, Kershaw, Darvish, and Lester, in every case it's the fastball being hit for the palpably highest BABIP.
 
Timing and control are obviously the crux of this issue. They'd have to be. To that point, I am not fully convinced that hitters are as prejudiced toward the heat as you make them out to be. 58% of pitches are fastballs, according to Baseball Info Solutions, and that encompasses every type—sinkers too. (Splitters constitute a disparate 1.6%.) Probably half the pitchers in baseball throw a plurality of breaking stuff, so I'm uncertain these "biases" are quite so pervasive. And some of the more successful pitchers in the game (Arrieta, Tanaka, Uehara, Cecil, just that I know of) could reasonably consider the fastball their #2.
 
Mariano Rivera is perhaps the most extraordinary pitcher of all time, and he threw one pitch.
 
If nothing else, I cling to the notion that pitch mix could be made more efficient. These are FanGraphs' anointed "pitch weights." By this measure fastballs in 2014 were worth negative-365 runs in aggregate. Curveballs cost pitchers nearly eighty, and all the rest were positive.