Four-Seam Fastballs and the Home Run Tidal Wave
David Laurila posted another of his typically excellent interviews this morning at Fangraphs, and it’s especially of interest for M’s fans. He interviewed OF prospect Gareth Morgan, the Canadian power prodigy I’ve talked about frequently here – the guy whose inability to make contact look like an absolutely fatal flaw, one that might end his career before he got to full-season ball, but who’s now having a nice spring for Clinton in the MWL…despite a K rate that’s down to “only” 39.3%. Asked what he’s doing differently now than in previous years, Morgan responded, “I’ve simplified my approach to where I’m just looking for a fastball to drive to the middle of the field.” You hear echoes of that from other players, including many big leaguers: you simply can’t adapt to everything, so many players simply look for a pitch – preferably a “fastball to drive” – in a specific zone, and swing like hell at it. Other pitches, even pitches that they know may be called strikes, they watch.
This got me to thinking about a throwaway line in my post about bullpens the other day. No, not the one about how the M’s should go to the bullpen rather than let Christian Bergman pitch 7 innings…that was just…forget about that. I was trying to figure out why batters were hitting more HRs and just generally hitting better against relievers this year, and wondered, “Maybe the fact that relievers throw more four-seam fastballs matters.” Put it together with Morgan’s quote above, and you’ve got something to look into. Two things, actually. First, are batters hitting more HRs off of four-seam fastballs than other kinds of pitches (and a higher percentage than in recent years)? And if so, does that mean that pitchers should throw different pitches?
Thus far this season, four-seamers have been hit for HRs at a greater rate than any other pitch. I don’t think this is all that shocking, but I’ll admit to some surprise at just how different contact on four-seamers is compared with other pitches. There have been 581 HRs on four-seamers, and 500 on every other type of fastball, plus off-speed pitches. Now, of course, we should control for swings and the number of pitches thrown. On a per-swing basis, we get the following:
| Pitch Type | HRs per Swing |
|---|---|
| 4-Sm | 2.00% |
| SI/2-Sm | 1.72% |
| SL | 1.58% |
| CH/SPL | 1.56% |
| CU | 1.37% |
No matter how you look at it, four-seamers are more likely to end up in the seats. The reason’s pretty clear, too: batters hit them harder and hit them in the air. Here’s another, very similar, table which looks at the average exit velocity off the bat and the launch angle by pitch type for 2017:
| Pitch Type | Exit Vel. | Launch Ang. |
|---|---|---|
| 4-Sm | 88.4 | 15.7 |
| SI/2-Sm | 88.2 | 6 |
| SL | 85.2 | 11.4 |
| CH/SPL | 84.6 | 7.6 |
| CU | 85.4 | 8.5 |
It’s no wonder batters hit proportionally more HRs on four-seamers than 2-seam/sinkers: their launch angle is about *10 degrees* higher! They hit the average four-seamer 3 MPH harder than curves, and about 4 MPH harder than offspeed pitches.
This year, batters share of HRs coming on four-seamers is up slightly compared to 2015, the last year before this current wave of HRs. Thus far, 41.26% of all HRs have come on 4-SM fastballs, compared to 39.69% in 2015. That’s a tiny change, but of course the percentages are multiplied against several thousand HRs. Of note: the real shift we’ve seen since 2015 is a *decline* in the share of NON-4-SM fastballs. The “all other fastballs” category accounted for 28% of all HRs in 2015, but 25% this year. All breaking balls plus knuckleballs have accounted for fractionally more HRs this year, but the sheer numbers are greatest with four-seamers. The easiest way to see the change is by looking at how HRs per swing has changed over the Statcast era:
| 4-SM | HRs per Swing |
|---|---|
| 2015 | 1.66% |
| 2016 | 1.83% |
| 2017 | 2.00% |
Remember, the denominator here is swings on four-seam fastballs, so the numbers really add up. There were well over 100,000 swings in 2015 and 2016. If you applied 2017’s HR/SW percentage to 2015’s number of swings, you’d have *400* more HRs hit, just on four-seamers. The entire HR surge of 2016 produced 700 (ok, 701) more HRs than 2015. The difference between 2016 and 2017’s rate would account for about 200 of those HRs. The fact that it’s still going up so much from 2016 to 2017 seems to highlight the degree to which batters are keying in on four-seamers, and that’s before we’ve even gotten to the warmer months, when HRs typically rise. However you look at it, a good chunk of this big wave of HRs washing over the league is the result of much better contact on four-seam fastballs.
So pitchers should throw sinkers instead, right? Well, we haven’t proved that. There are plenty of other reasons to throw four-seamers, like a higher whiff rate and potentially a lower BABIP. There’s a reason why teams like the Tigers, Red Sox and Nationals are throwing more four-seamers and throwing more of them up in the zone. Pitchers like Justin Verlander and Max Scherzer are great not in spite of their four-seam usage, but because of it. Both of them are reliably among the leaders in HRs-allowed, but this doesn’t make them ineffective.
Baseball’s strikeout rate has been rising for a long time, and it’s at least in part due to the fact that teams are no longer scared off of high strikeout totals. Once teams stopped selecting for contact and started selecting for a broader measure of offense, the strikeout rate was bound to go up. Teams have also selected for strikeout pitchers, and thus the trend got another boost. As strikeouts rose, a player with a high K rate was no longer freakishly abnormal. What’s this have to do with HRs? I wonder if we’re at a point where teams will look past a high HR rate if a pitcher’s overall effectiveness looks decent. Obviously, it’s harder to do this with relievers, as old friend Shawn Kelley’s now discovering. He’s given up 5 HRs on four-seamers this year, and even with a positive WPA and a sky-high K rate, he’s losing his closer job in Washington, DC. But Max Scherzer? While he gives up plenty of HRs, he’s still one of the better starters in the National League.
The best of both worlds, of course, would be the benefits of the four-seamer without its loud side effect. This was Chris Young’s M.O. for some time, and it’s worked for others here and there. Only a few have had the command necessary to pull it off for a long time (longer than a season or two). Chris Young’s trick was to locate the ball very precisely up in the one to make batters hit underneath the ball. He could do it (until he couldn’t), but while high pitches have lower SLG%-against in recent years than low fastballs, it’s not like throwing high FBs is some sort of HR-prevention trick: a higher percentage of HIGH four-seamers has been converted into souvenirs than low ones.
Knowing this, and I’m sure teams do, there are two possible responses. One is to all but abandon the four-seam, especially high ones, and just throw a million low sinkers or a blizzard of breaking balls. The Houston Astros exemplify this, as Dallas Keuchel and company lead the majors both in the percentage of low pitches thrown, and they’re throwing a ton of breaking balls+change-ups, too. More interesting in light of all of this is the Colorado Rockies. They play at altitude, and batters have long hit more HRs in Denver than elsewhere. Breaking balls don’t break the same, etc. It’s no surprise then that they’ve emphasized ground ball contact; they were 3rd in MLB in GB/FB ratio in 2015, 2nd in 2016, and are #1 (easily) this year. What’s more interesting is that they’ve taken the opposite approach to the Astros. In 2015, the Rockies threw four-seamers 36.6% of the time, 12th highest in the league, but far, far below the Rays, who were nearly at 50% (49.34%). The following year, the Rockies moved up to 4th in four-seam percentage, clocking in at around 43%. This year, they’re #1 by over 2 full percentage points. The numbers have all come down as fastball usage overall drops, but the Rockies are still climbing: they’re at 46.6% four-seamers, and lead the league in swings and balls in play off of four-seamers. Given everything we’ve talked about, you’d expect the Rockies are playing 2000-style baseball again, but they’re not. Their HR rate is much lower than average, and waaaaay below the Mariners’. How’s this possible? I don’t know, I’m sorry…I’ve let you down again. What I can say is that they’re getting the lowest average launch angle on four-seamers, and again, it’s not particularly close. They’re giving up well-below-average exit velocities on them as well. They don’t appear to be pitching up, or down, or anything, as you can see from this heatmap showing all four-seamers put in play (it looks pretty similar to the heatmap of all four-seamers thrown, too):
I get what the Astros are doing, and why they’re doing it. I don’t quite understand what the Rockies are doing, or rather, why what Colorado’s doing seems to be working, and I would really, really like to know. In summary, four-seamers are going for HRs more these days, and that’s part of the reason for the HR barrage league wide. It seems like batters are waiting for specific pitches and then jumping on them, ala Gareth Morgan’s quote to David Laurila. This explains why HRs are going up along with K rate, and potentially why relievers (who throw more four-seamers) are giving up more runs. It’s not necessarily the case that teams should just swap out four-seamers for sinkers, though, as teams like the Nationals and Rockies illustrate.
Comments
2 Responses to “Four-Seam Fastballs and the Home Run Tidal Wave”
Leave a Reply
You must be logged in to post a comment.
Here’s a thought. The pitch is a four-seamer. Does the fact that it is showing four seams make it easier to pick up? If a hitter is looking a fastball in a certain area, is that easier to do if the pitch is a four seamer?
That’s my thought, Steve. As a four seamer’s the most common FB type, its movement is the most ‘traditional’ or natural to a hitter. If you see a pitch in your zone, you need to anticipate where it’ll end up a split second later, when your bat actually contacts it. I think it’s easier for batters to do that instinctive bit of extrapolation with a four-seamer’s movement compared to a sinker or anything else.