Investigating Fastball Velo Changes
How often does a pitcher gain or lose fastball velo? Are there certain types of pitchers that are more likely gain or lose? I answer these questions and more using 2021-2023 Statcast data.
It is that time of year. Mid-February. Pitchers and catchers are reporting, soon we will have Spring Training games being played. We will soon be hearing about players who are injured due to pickup basketball injuries or because they fell down the stairs. We will be hearing about players who lost weight, pitchers adding new pitches! But everybody’s favorite, I think, are the reports about pitchers who have gained velocity.
We’ll get some actual data on this when games start being played. A bunch of these Spring Training stadiums have Statcast infrastructure, so I will certainly be scraping and analyzing that data as soon as it is available.
But we don’t even have to wait that long, we have already heard about Shane Bieber having added velocity back. Those reports came from his trip to Driveline.
So the table has begun being set, but I would like to add some candles to the centerpiece. Let’s talk about the general topic of pitchers adding velocity.
Does Velocity Matter?
Yes.
That was easy enough!
How Often Does It Happen?
I’m looking at 2021, 2022, and 2023 data and limiting to only 4-Seamers and Sinkers. I require at least 50 pitches thrown, and then take the average velo of each pitch from each pitcher.
We have 1,163 data points here. Meaning 1,163 different pitches to study. Of those 1,163, there are only 30 instances of a pitcher adding at least two miles per hour to his fastball from one year to the next That is just 2.5% of the time.
If we look for instances where a pitcher added at least one mile per hour to their fastball, that happened 12.1% of the time.
There were 14 times when a pitch lost more than two miles per hour, that’s just 1.2% of the time, and there were 191 times when a pitcher lost at least one mile per hour (16.4% of the time).
So it’s more common to lose velo than to gain it. Surveying all of the data points, the average change was a loss of 1.15 miles per hour.
This makes inherent sense to me. You usually have to be at the top of your game to make it into the Major Leagues. Not many developing pitchers are promoted to the Majors. But since a veteran player cannot be sent to the minors, and they often have multi-year contracts, teams tend to let guys still pitch for them even as they are on the downslope of their career. Adam Wainwright for example. The Cardinals let him pitch at least a couple of years after he was clearly rapidly descending physically, so that velocity drop was allowed to make it into the data. Meanwhile, if a 19 year old kid adds a bunch of velo, that will usually happen in the minors leagues.
Where Are Velo Gains Most Common?
First, let’s look at age. I took each player in my data and merged in their age at the time of that season, and then averaged out the velo changed by each age group.
The -0.5 average on the 23 year olds is a result of a small sample. That 23 year old bucket meant that a pitcher threw at least 50 fastballs as a 22 year old and a 23 year old, which is quite young. My data is also only going back to 2021. There were only 12 pitchers studied here, so it’s not a reliable sample size.
We also only had seven seasons to study for players over 40, and four of those seasons were Adam Wainwright (two for each fastball variation he threw in each year sample).
Rich Hill was able to add a bit of velo to the four-seamer in 2022, but he lost most of that gain in 2023. Darren O’Day went out and pushed his very slow fastball up a half of a mile per hour in 2022. If we had more data to study, I am quite certain that the general trend of losing lots of velo with age would be stark.
So the bulk of the data is on pitchers between 26 and 32.
We see velo gains most often in the younger years. Ages 24-26 are the only ones where the average change was upward.
The biggest drop was at age 32, and the drops get less severe from 33-36. This is probably survivorship bias. Meaning that the pitchers that pitch well into their thirties are going to be guys that don’t lose velocity. If you’re 33 and losing velocity, teams aren’t going to want you on their Major League team. So the velo decliners get filtered out automatically.
There are a lot of statistical things to consider here, but it seems clear that we should only have any sort of expectation of velocity gains from pitchers under 27 years old.
Next, let’s look at “velocity bins”. It would stand to reason that a pitcher throwing 90 has a greater chance to add velocity than a pitcher already throwing 98. So I took each year one average velocity and rounded it to the nearest whole number, and then we can study by these “bins”.
The assumption is proven correct. Velocity gains are much more common from players with low beginning velocities. The same sample size stuff applies, there are very few pitchers in the league average less than 90 on a fastball (just 60 to be exact) or more than 98 (16 to be exact). So that big +2.38 for in the 84 bin is because only one pitcher averaged 84 on their fastball and then pitched the next year as well.
The bulk of the data here is between 92 and 97 mph. It would seem that it’s somewhat easy to add some velocity if your fastball starts at 92. That’s the only bin above 91mph with a positive average change.
There are 127 pitchers in the data in the 92mph bin. 68 of them increased their velo the next year (54%), leaving 58 (46%) that decreased.
But still, only 16 of them added more than one mile per hour (12.5%), so it’s not like it was common to see significant velo added. Only five pitchers starting at 92mph lost more than one mile per hour (3.9%).
Does Lost Velo Come Back?
This is question concerning Shane Bieber this year. It would seem more likely for a pitcher that at one point had more velocity and then lost it to gain it back.
The Biebs averaged 92.75 on his four-seamer in 2021. He lost 1.49mph in 2022, and then added only 0.08 back to it in 2023.
We should probably have more data to do this right, but I’m plunging forward anyways. What I’m looing for now are pitchers that qualified in all three seasons and lost velo in 2022. Then we’ll study what happened in 2023 to see how often the velo comes back.
I found 134 pitchers that lost at least 0.5mph on their fastball in 2022 and then threw at least 50 more of those fastballs in 2023.
I find that 29 of them added at least that 0.5mph back in 2023 (22%), and 46 of them lost at least another 0.5mph in 2023 (34%). Not a good sign. It seems following a year of lost velocity, you are more likely to lose even more the following year than you are to gain it back.
Let’s look back at just the pitches under 93 to start with though. So now I’m taking
Pitchers that average under 93mph in 2021
Lost at least 0.5mph in 2022
The average change from 2022-2023 in this group was a lost of 0.05 mph. Meaning that mostly things stayed the same. 13 of the 44 names (29.5%) lost at least another 0.5mph in 2023, but 14 (32%) added at least 0.5mph back.
So, when we take out those high-velo pitchers, you are about equally as likely to add or lose more velo following a year where you lost it, but by far the most likely scenario is that your velo comes in very close to the previous year.
Checking the correlations:
There are high correlations here, because pitch velo does not change much at all over even a three-year sample, but you can see there that year three is better predicted by year two than by year one. That’s pretty obvious, but now the obviousness has some numbers on it.
Conclusion
So what does all of this mean? Here are my main takeaways in order of importance.
Velocity changes are rare
It’s more common to see lost velocity than gained
It’s more common to see velocity gains from
Pitchers under 26
Pitchers averaging under 93mph
Once it’s lost, it usually doesn’t come back
But you’re more likely to get it back if you started under 94mph (this is basically the same point as 3b).
But the good news is with Bieber, we don’t have to guess for long. We’re guessing right now, but we’ll know with a good degree of certainty where he’s at once we see him make a spring training start or two. If he comes out and throws 50 four-seamer and averages 93 on them, then that average should be our new expectation. If he’s still under 92 several starts into spring, all hope is lost.