New Data in Baseball Aging Curves and the Toronto Blue Jays

 

After reading an incredibly insightful piece by Jeff Zimmerman at FanGraphs, I was inspired to write an article about the Blue Jays’ own young(ish) players.

 

 

First, I urge you all to read the article to which I’ve linked above. It’s a terrific look at how Zimmerman has observed the aging curves of batters has changed, particularly since 2006 when more severe punishments for PEDs have been put into effect.

 

The old idea of an aging curve was that players would start their careers at a certain point of playing ability in their rookie seasons, improve for a few years and then decline with age. In the past seven years, however, the curve doesn’t rise to that peak production before declining. Players actually start at their peaks and we can only expect to see decline, starting around their age 25 or age 26 season.

 

Now, Zimmerman uses two different measures of offensive production, wOBA (Weighted On Base Average) and wRC+ (Weighted Runs Created Plus) and gets two different results (click on the links for a thorough explanation of both). The first, wOBA shows a constant decline, right from the get-go in the past seven years (since 2006) when compared to the ten-year period previous (from 1995 to 2005). Zimmerman attributes part of this steady and immediate decline in the past seven years to the fact that wOBA across the league has been dropping fairly steadily since 2006.

 

Zimmerman ran the tests again using wRC+ to find a slightly different result. Because wRC+ is already weighted towards league average, it eliminates natural drop that you get when the lower-scoring environments of the past seven seasons are factored in. What Zimmerman sees is a constant plateau from the Age-21 to the Age-25 seasons with a steady decline thereafter. In other words, once a player hits about 26 years old, his offensive production will start to drop relative to league average.

 

Zimmerman hypothesizes that players no longer improve once they get into the league because, for the most part, they’re better prepared for the major leagues by the time they get there due to much more investment in player development programs by teams.

 

Now, when looking at overall trends, it’s clear that you need to look at large data sets and the aging curves of individual players are going to conform much less to this overall picture. I thought I would look at some Blue Jays players and how their careers have looked with these things in mind.

 

I plotted the returning Blue Jays’ hitters wRC+ against the years in which they achieved them. What I got was a pretty jumbled graph but what we can see is that Zimmerman’s theory of a slowly descending performance after about Age-26 is debunked when we look at individual players. The last two selections on the graph are the results of two different projection systems (Steamer and Oliver) that are available on Fangraphs.

 

 

When we dig into what’s going on here a bit more we realize that there are things that are unaccounted for in a large-scale data analysis like the one that Zimmerman undertook. We don’t account for either of Jose Bautista‘s or Edwin Encarnacion‘s later-career surges: for both it was in their Age-29 seasons.

 

Several players have large drops after their rookie seasons (partial seasons for some). These players include Jose Bautista, Brett Lawrie, Jose Reyes, Adam Lind and Josh Thole. Some players improved after their rookie seasons like Maicer Izturis, Edwin Encarnacion and Colby Rasmus, only to have their performances drop off again.

 

Basically, when we look at our chart, it’s pretty much a crapshoot trying to predict players’ actual performances based on overall expectations. With young players like Brett Lawrie and Anthony Gose, that problem compounds itself with the fact that they’re both supposedly in their peak period (pre-Age-26) and yet, we haven’t seen anything close to what we expect to be their peak performances. Will they learn and allow things to “click” they way they did later in their careers for Encarnacion and Bautista? Or will Lawrie and Gose continue to give us what they’ve been giving us for the next few years before starting to decline?

 

If we look at Rasmus, this trend obviously doesn’t explain the deep trough between the two peaks of 2010 and 2013. With Rasmus entering his Age-27 season in 2014, are we supposed expect his decline to start now?

 

Obviously, this doesn’t nullify the work that Zimmerman has done over at Fangraphs. What it shows is that you can’t exactly expect the data to be true for all players when you look at individuals. In fact, the individual details don’t correspond at all to Zimmerman’s large data sets.

One thought on “New Data in Baseball Aging Curves and the Toronto Blue Jays

Comments are closed.