Quick And Dirty Aging Curves With Exit Velocity.

Two weeks ago I wrote about what I call “Peak Angles” for batters. That is the angle at which they produce the highest ratio of high exit velocity to low exit velocity batted balls. The working theory is that this angle roughly correlates to swing plane, and when you know swing plane you might be able to work out valuable information. For example, you might be able to gain insight on batter versus pitcher match-ups by comparing the average pitch plane to the average bat plane. There are other factors involved, obviously, like contact skill and plate discipline, but swing plane could prove to be valuable in the long haul.

Over the past two weeks this line of thinking brought me to two different lines of inquiry. First, can you use the “Peak Angles Exit Velocity” to predict the maximum effective angle for a batter? Second, how do the most valuable launch angles and exit velocities age over time?

Well, I worked on the first question a bit, and it turns out to be a bit more difficult than I first imagined, so I pivoted to the second question. I have very preliminary findings to show you. This should all be double checked with a more vigorous approach, but hopefully I’m in the right ballpark.

I have created the following chart, which depicts wOBA with respect to launch angle and exit velocity.

This chart shows Vertical Launch Angle on the X axis and Exit Velocity on the Y axis. Everything on the left of the chart is a ground ball, on the right are fly balls. The top are hard hit, the bottom are weakly hit. Everything blue/purple has little value, and everything yellow/orange/red has high value.  Do you see that dark red area at the top of the chart? Let’s zoom in on that.
This region spans from 0 degrees to 40 degrees, and from 95 miles per hour all the way up to max velocity. I’m going to call this a Nitro Zone. It is similar in some ways to the “barrels” stat, although not exactly the same.

This area is a high value zone. Balls hit in this area have a .645 batting average, 1.369 slugging, and .834 wOBA. Of these balls, 17% are home runs, 18% are doubles, 28% are singles, and 2% are triples. Batters who produce a large number of balls in this area will be very successful, but we’re curious how batters may age. From the onset, we might expect the velocity factor will matter more than the launch angle factor.

I counted how many balls each batter hit into this area in each of the past three seasons. Below you can see a chart showing the Year 1 to Year 2 changes for each. Generally speaking, it appears to be a mostly linear relationship, although not quite. This trendline has an R squared value of .635. So that is fun, it correlates pretty well from one year to the next. Now to get into the age component. I grouped the players by age and found the difference in balls hit into the Nitro Zone from Year 1 to 2. So, for example, year 33 to 34, or 24 to 25. I found the mean, 95th percentile, and 5th percentile changes between each ages which you can see in the following chart. The top set of points in red show the 95th percentile, the middle group in green show the mean, and the bottom group in blue show the 5th percentile.


Note the changes in the mean. The curve has a steep fall off after the age 21-23 seasons, stays pretty steady for a while, and then drops off around age 33. By age 35 it is dropping pretty rapidly.

The 95th and 5th percentiles should serve to establish the upper and lower bounds you might see from any given player from one year to the next. Note that the variance is largest in the primary playing years, from age 24 to 30, then closes up a bit in the early 30s. That is likely caused by survivor bias.

So, what is there to learn from this? Well, this aging curve doesn’t appear much different from most other aging curves you have come across in the past, but maybe it serves to help you determine the risk associated with certain aging players in today’s market.

Keep in mind that MLB batters have hit about 23.7% of their balls into this Nitro Area, in case you want to do similar calculations on your own using Baseball Savant.

Earlier in the off season I put together some aging curves for exit velocity, which you can view below. For this analysis I have split batters into two categories. First, batters who have “High” exit velocity, which I define as those with an average over 88mph. Second, batters who average below 88mph, labeled “Low.”
Notice how those with a “low” exit velocity group gains exit velocity through certain parts of their aging cycle, namely during the early part of their career. This could be an issue with small sample size throwing off weird results. It could also be superior coaching, scouting, and a change of approach that leads to better results over time. Eventually, the exit velocity begins to fall, and by age 27 players seem to have diminishing returns. Then we see a bump that is likely caused by survivor bias during the early 30s.

The high exit velocity group has a totally different aging curve. It is down, down, down. High exit velocity guys appear to, on average, lose velocity each season they play. You see a large drop after the initial few seasons, which, again, I believe is linked to superior coaching. Young batters may be swinging too hard, which can lead to issues with poor quality batted balls due to hitting the wrong pitch, or increased swing and miss rates. Perhaps young batters are coached to reign it in a little, which we can see as a dip in exit velocity during their second and third major league seasons. Or, perhaps, it is a sample size issue.

Either way, these two exit velocity aging curves seem to match up with the “nitro” aging chart from above. Namely, all three appear to draw the line around age 34. After that point, batters seem to age very poorly. Even leading up to that point, around age 32 or 33 you tend to see potential issues. Some batters see small incremental improvements in their age 33 and 34 seasons, followed by sharp declines.

You might want to look out for: Miguel Cabrera, Robinson Cano, Ian Kinsler, Curtis Granderson, and Yadier Molina. Dustin Pedroia, Hunter Pence, Joey Votto, and Kendrys Morales each saw declines in this “Nitro” zone batted ball figure this past season, and all three are age 33.

Votto saw a large decline during his age 32 season as well. He has fallen from 38.1% at age 31 to 30.8% at age 33. The MLB average is 24%. He has also had a large decline in general exit velocity, falling from 90.4 to 87.2 mph from 2015 to 2017. Of course, Votto has manage great production over these three seasons, but there has to be a breaking point here somewhere.

At the end of the day, both of these approaches to aging curves are likely explaining the same exit velocity phenomenon, but it is interesting and perhaps more valuable to measure it using two different methods while finding very similar results for each.

The “Nitro” method measures the number of balls a batter produces in the highest value section of the launch angle/exit velocity space. Dips in these numbers are generally bad, since the lower exit velocity area immediately beneath this “Nitro zone” is the “Dead Zone.” Those balls have basically zero value.

The second method is strictly looking at average exit velocity, with no other stipulations.

In both cases players appear to hold roughly constant exit velocity between ages 26 and 30, and in both cases the age 33-34 seasons are the last time improvement can be expected from a player. Both point towards peaks in velocity in the initial seasons for young players, followed by sharp decline into their primary playing years. Further research should be done into all rookie players, perhaps all players see a sharp decline in exit velocity after their first major league season.





Andrew Perpetua is the creator of CitiFieldHR.com and xStats.org, and plays around with Statcast data for fun. Follow him on Twitter @AndrewPerpetua.

18 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
DBA455
6 years ago

This is really interesting, differentiated stuff.

Allow me to preemptively congratulate you for being hired by a club and taken out of the public sector some time in the next 12 months!

johnnycuffmember
6 years ago
Reply to  DBA455

seriously though. it’s gonna kill my fantasy team when xStats goes private.

sgnthlr85
6 years ago
Reply to  johnnycuff

Is this really happening? If so, when? I’m currently unable to access the baseball savant webpage.