I don’t know if you just watched the Royals and the Tigers. Probably not. Neither team is very good. The Royals won 1-0. Jakob Junis was terrific. The teams combined for just eight hits, and seven of them were singles. Somewhat more notably, the teams combined for a ground-ball rate of 32%. The average launch angle in the game, according to Baseball Savant, was a hair over 22 degrees. What does it mean? By itself, hardly anything. The game, however, was not by itself.
Nothing that follows ought to come as much of a surprise. We’ve been talking about air balls and launch angle for a couple of years. It’s not just fan sentiment, either; we know that more and more players are buying in. And, based on the data from this most recent spring training, it seemed as if something was going to happen. More balls were being hit in the air in the spring, suggesting we could see the same when the games started to count. There have been hints. We know that hitters are less fond of grounders than ever.
Even so, I didn’t expect quite the numbers we have. You don’t need to remind me of how early it is. Yet, already, there’s evidence of a significant batted-ball shift.
We’re into our fourth season of Statcast data, right? And the stuff everyone loves are exit velocity and launch angle. Check out these league-average launch angles. You might notice a pattern.
- 2015: 10.1 degrees
- 2016: 10.8
- 2017: 11.1
- 2018: 13.3
Between 2015 and 2017, there was a league-wide increase of one degree. Already, we’re seeing a sign of a league-wide increase of more than two degrees. Yes, of course, nothing can be confirmed until we have a lot more information, but this is the kind of thing you’d expect to fairly rapidly find its level. This is the sixth day of the 2018 regular season. We can break launch angle down on a day-to-day basis!
- March 29: 12.5 degrees
- March 30: 13.0
- March 31: 13.5
- April 1: 14.8
- April 2: 12.0
- April 3: 22.4 (incomplete)
Five straight days of 12 degrees or higher. The average for all of 2017 was 11.1. And while any single-day sample is small, each of the first five days had well more than 600 batted balls. On the league-wide scale, numbers aren’t quite as volatile as you’d think. On any given day, there’s an awful lot of baseball being played, and that helps to smooth out some of the randomness.
Launch angle is one way to attack this, but it’s also limiting. We just don’t have that much of a Statcast history. And, I don’t know, maybe you aren’t such a fan of the batted-ball calibration, or whatever. Maybe you don’t trust that sort of technology. Similar to launch angle, we have ground-ball rate. Here are league-average ground-ball rates going back to 2008. The y-axis is truncated, but I’m not trying to deceive you. That’s why I’m including this note that the y-axis is truncated.
Huge drop. Last season, 44.2% of batted balls were grounders. In the early going this season, 40.7% of batted balls have been grounders. It’s not like grounders have completely disappeared, and if there were a given hitter who experienced this sort of shift, you wouldn’t even call him that different. But this isn’t about one hitter. This is everyone, combined. It’s difficult for a league-average number to change by very much, but that’s what we might be seeing.
You’re absolutely right to point out that it’s not fair to compare 2018 to previous full seasons. 2018 is literally just a few days old. Thankfully, our splits leaderboard can come in handy here. There have already been more than 3,000 batted balls. I pulled up the splits leaderboard and examined the league on a week-to-week basis going back to 2008, including only weeks with at least 2,000 batted balls. That query yielded 263 weeks of major-league play in all. This plot shows weekly ground-ball rates.
This, I think, is the strongest evidence that something is changing. That is, aside from all the anecdotal evidence we already had. We’re looking at the lowest ground-ball rate in a week in at least a decade. The current rate is 40.6%. The lowest rate in any week last season was 42.3%. That was the lowest rate in a week since July of 2010. And the further back you go with this, the more you run into changes in how grounders, liners, and flies were recorded. Definitely, just considering the last handful of years, we’re seeing something we haven’t seen. It could be random noise, and we’ll find that out eventually, but I think that idea is unlikely to be true.
In many ways, it’s absurd to look at any league averages so early on. We’re not even at a point where every starting pitcher has started a game. I wouldn’t have written this post at all had that weekly analysis not turned out as it did. That’s what sold me that we’re probably seeing a change. As a side note, there’s a consequence of this: we’re also looking at our second-highest weekly infield-fly rate since 2008. Batters can try to hit the ball in the air. Not every air ball is a winner.
To repeat what I noted, this isn’t a shock. You could argue the trend started a short while ago, when the ball started acting more lively, so this wouldn’t even be anything new. And maybe it’s the sort of thing that would snowball — as hitters try to lift pitches down, pitchers will throw more pitches up, and those pitches seldom get hit on the ground. So the ground-ball rate would over-shift, until hitters level out again. A lot of the trends we see end up being cyclical. Baseball is a never-ending tug-of-war. Hitters and pitchers are forever just trying to figure one another out.
It’s just, we’ve talked about air balls, but we haven’t yet observed a big change in the grounder rate. Based on early indications, the change is underway. Part of it is that certain hitters have changed their approaches. Part of it is that teams are selecting hitters with certain approaches. The air balls already look like they’re here. The statement itself is not a surprise. That we can make it so early in 2018 is what’s more surprising.
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.