Note: The attrition data and chart was updated, showing a larger gap for just about all age cohorts
On May 26, the Twins released 33-year-old starter Jason Marquis. Considering how bad Twins pitchers have been this season, it really spoke to how bad Marquis’ numbers were to that point in the season.
In seven starts, Marquis posted a 8.47 ERA and a 7.25 FIP, numbers driven largely by a 7.5% strikeout rate, a 27.3% HR/FB and a sky-high .352 BABIP against. There was some speculation that, since Marquis has lost about 2 mph on his fastball since 2009, it was likely that he’d become more hittable — particularly at age 33.
This led to some discussion on Twitter about whether there was a greater likelihood that Marquis’ velocity drop at his age was more of an issue than if he had been a younger pitcher. It’s a legitimate question — whether diminished velocity has an impact, depending on age.
The answer seems straight-forward: calculate the average performance for each age cohort in the season they lost at least 1 mph and for the season after they lost the velocity, then compare the two. Easy enough. But the answer is complicated because some pitchers will have left the sample by the following year.
First, let’s take a look at how velocity loss impacts performance, by age. I gathered data on all pitchers between 2002 and 2011 who threw at least 40 innings each year. I then calculated weighted averages for change in ERA- and FIP-, K%, BB% and fastball velocity (FBv) for each age cohort for those pitchers who lost at least 1 mph of fastball velocity. Here are the results:
In just about every cohort, pitchers performed worse as their velocity declined. In some cases, pitchers improved their walk rates — but in terms of their adjusted ERA’s and FIP’s — their performances were worse.
Generally speaking, the loss of velocity impacts older pitchers more in terms of their adjusted FIP than for younger pitchers.
What about the year after the drop?
Looking at pitchers who had three consecutive years in the data set, we find this:
Here, the story is more mixed. In some cases, pitchers improved their FIP- — or at least held the line — largely due to an increase in their strikeout rates. Interestingly, in many cases, their fastball velocity either held steady or actually improved a tick from the previous year.
We might be tempted to say that losing velocity is not the end of the world, as many pitchers rebounded well even as they aged. The problem is that the data on performance only includes those pitchers who managed to throw at least 40 innings the following year. A significant number of pitchers dropped out of the sample*, and so we have a survivor issue to deal with.
To calculate the attrition rates, I took the same sample and derived the overall attrition rate for any pitcher between Year 1 and Year 2 — depending on their age. Next, I calculated the attrition rates among those pitchers who lost at least 1 mph in Year 2. The rates are compared in the graphic below:
Not surprisingly, losing velocity does not help a pitcher’s chances of throwing 40 innings the following year, though there seems to be a significant spike in the impact between 25 years-old and 27 years-old, and then again from 34 years-old through 39 years-old. The samples get smaller at both ends of the age spectrum, but there appears to be something to the notion that losing velocity in later years is more harmful for pitchers.
Why is this the case? Here are a few potential reasons, which are mostly intuitive:
– It’s less likely that a team is going to take a chance on an older pitcher who recently lost velocity, compared to a younger arm. Wasting a roster spot on a 34-year-old who you hope will recover isn’t as appealing as a younger player.
– Velocity drops later in pitchers’ careers lead to lower overall velocity than the same drop for younger pitchers. The average fastball for 24-year-olds in the sample is 91.1 mph, compared to 89.4 for 34-year-olds. And, remember, that’s the 34-year-olds who lasted in the sample.
– Players become more injury prone as they age, and recovery becomes more difficult. It’s likely that, as a pitcher hits his mid-30s, velocity decline is due to injury — and recovering from that injury becomes more difficult.
*Not every pitcher missing from the data set in the following year was necessarily out of the league. In some cases, they may have simply pitched fewer than 40 innings. However, if they where throwing fewer than 40 innings, it likely means they were either injured or ineffective enough to warrant a low workload.
Bill leads Predictive Modeling and Data Science consulting at Gallup. In his free time, he writes for The Hardball Times, speaks about baseball research and analytics, has consulted for a Major League Baseball team, and has appeared on MLB Network's Clubhouse Confidential as well as several MLB-produced documentaries. He is also the creator of the baseballr package for the R programming language. Along with Jeff Zimmerman, he won the 2013 SABR Analytics Research Award for Contemporary Analysis. Follow him on Twitter @BillPetti.