Where Have the Fastballs Gone Missing?

Early Thursday, I listened to an exchange between Buster Olney and Indians president Chris Antonetti. As you’d expect, there was talk about the Indians’ winning streak, and about the impressive play of Jose Ramirez. But Olney also asked Antonetti about an observation that had been relayed to him by some number of league evaluators. In the opinions of those evaluators, one area where the Indians stand out is in their reluctance to throw predictable fastballs. Pitchers have been taught forever that the fastball needs to be established early on. What if a team simply didn’t believe that?

Listening to the segment got some gears whirring. This isn’t a post about the Indians. This is a post inspired by an observation about the Indians. Let’s have a little talk about fastball usage.

It’s been observed that fastball usage, overall, is down. It’s not down in such a way that the current version of baseball is unfamiliar, but a trend is a trend. You don’t need to know very much in order to be able to interpret the following.

Teams aren’t using so many fastballs. Compared to a decade ago, the league-average rate has dropped five percentage points. I just wrote a separate post touching on this. In that post, I separated four-seam fastballs and sinking fastballs, in order to show that four-seamer usage has held mostly stable. It’s mostly sinkers that have been going away. Whether permanent or temporary, that’s what’s been going on.

I wanted to come at this from another angle, so I sat and made laborious use of Baseball Savant. Forget about separating fastball types. For this analysis, I blended them again. I wanted to look at how fastball usage has changed over the years, by count. We have 10 years of information, and there are 12 different counts. So I gathered 120 individual data points. Here they all are, arranged in a table not a single one of you is going to take the time to consume. (I wouldn’t try to consume it, either. I’m just trying to be transparent.)

Fastball Rates By Count
Year 0 – 0 0 – 1 0 – 2 1 – 0 1 – 1 1 – 2 2 – 0 2 – 1 2 – 2 3 – 0 3 – 1 3 – 2
2008 62% 51% 48% 62% 52% 45% 72% 63% 49% 77% 79% 63%
2009 64% 51% 48% 64% 52% 45% 74% 64% 50% 80% 80% 64%
2010 64% 50% 48% 63% 51% 44% 74% 63% 49% 80% 80% 64%
2011 63% 50% 47% 62% 50% 44% 72% 62% 48% 80% 78% 63%
2012 63% 51% 47% 62% 50% 44% 73% 62% 48% 82% 79% 62%
2013 63% 51% 48% 61% 51% 44% 73% 61% 48% 81% 79% 62%
2014 62% 51% 48% 61% 51% 44% 72% 61% 48% 80% 79% 63%
2015 62% 51% 48% 61% 51% 45% 72% 62% 49% 81% 78% 63%
2016 61% 51% 48% 60% 51% 45% 71% 59% 49% 82% 78% 62%
2017 60% 50% 46% 57% 49% 44% 69% 58% 47% 82% 76% 61%
SOURCE: Baseball Savant

That’s overwhelming, but, thankfully, we can simplify. We have the 12 counts, and we have the 12 numbers for this current season. I then calculated, for each count, the 10-year average. The last step was dividing the 2017 fastball rate by the 10-year-average fastball rate. This way, we can see from where the missing fastballs have been plucked.

Fastball Rates By Count
Count 2017 FA% 10yr FA% 2017 vs. 10yr
1 – 0 57% 61% 93%
2 – 1 58% 62% 94%
0 – 0 60% 62% 96%
2 – 0 69% 72% 96%
3 – 1 76% 79% 97%
1 – 1 49% 51% 97%
0 – 2 46% 48% 97%
2 – 2 47% 49% 98%
3 – 2 61% 63% 98%
0 – 1 50% 51% 99%
1 – 2 44% 44% 100%
3 – 0 82% 81% 102%
Overall 55% 57% 96%
SOURCE: Baseball Savant

This is the fun stuff. At least, to me, this is the fun stuff. Skip right ahead to the final row. What that shows is that this year’s overall fastball rate is 96% of the 10-year-average fastball rate. If you’d like to think of it in other terms, this year’s overall fastball rate is lower than the 10-year average by four percentage points. As we go by count, do we see an even distribution of missing fastballs? We most certainly don’t. We see that certain fastballs have been reduced this year more than others.

And there’s a pattern there attempting to emerge. The higher the position in the table, the greater the relative fastball reduction. The upper part of the table includes mostly hitter-friendly counts. Those are, traditionally, fastball counts. The lower part of the table includes mostly pitcher-friendly counts. Those are, traditionally, non-fastball counts. Nearly every count has seen some degree of fastball reduction, but the changes aren’t equivalent. Pitchers have reduced their fastballs more in counts in which fastballs would be expected. Think of this as a slight league-wide trend toward pitching backwards.

What is pitching backwards, anyway? It can and has been used as a pejorative. It’s been used to describe pitchers who don’t lean heavily on their heat. But what it’s really about is pitching in such a way that you avoid being predictable in predictable situations. That’s what I figure this is about. For just about every single pitcher, they’ve got better command of the fastball than anything else, but hitters like to hunt for fastballs when they can. So the fastballs can get drilled. Pitchers are slowly learning to keep things mixed up. They’re throwing fewer fastballs in general, and they’re especially throwing fewer fastballs in fastball counts.

Only one count has seen a recent increase in fastballs. That would be 3-and-0, where current fastball usage is higher than the 10-year average by two percentage points. It might seem counter-intuitive — what’s a better fastball count than 3-and-0? — but I think it makes plenty of sense. In many 3-and-0 counts, the pitcher basically gets a free pitch. Certain hitters are known to be particularly aggressive, but the league overall swings on 3-and-0 just 9% of the time. One swing out of every 11 pitches, in other words. It’s an opportunity for the pitcher to try to get a strike on the board, and so it follows that fastball usage wouldn’t drop, as it has in other counts. I’ll note that the 3-and-0 swing rate is very slightly up. This might be a partial response to the fastballs, as baseball searches for a very specific kind of equilibrium. There are so many areas where the game searches for balance.

Anyway, that’s all I’ve got. There hasn’t been anything groundbreaking. But the trends that are revealing themselves probably aren’t finished developing. Players now have a more full understanding of baseball than ever before. They know, more than ever, where they might be exploited. The last thing you want to be as a baseball player — at least, as a baseball pitcher — is predictable. Pitchers are gradually taking that to heart. There is still such a thing as the fastball count, yet it doesn’t mean quite what it used to. Nor, optimally, should it.

We hoped you liked reading Where Have the Fastballs Gone Missing? by Jeff Sullivan!

Please support FanGraphs by becoming a member. We publish thousands of articles a year, host multiple podcasts, and have an ever growing database of baseball stats.

FanGraphs does not have a paywall. With your membership, we can continue to offer the content you've come to rely on and add to our unique baseball coverage.

Support FanGraphs




Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.

newest oldest most voted
timmer
Member
timmer

Hi Jeff, I notice that you have the swing rate % for the league on 3-0 counts in the article above, can you tell me where I might find this information? I’m extremely curious about 3-2 swing rates, league wide and by individual batter, and if you could point me to where I could find it I would appreciate it. Thanks!