Aging Braun: An Expansion

Among Brewers fans, there is perhaps no player more divisive than Ryan Braun. As he should be: Braun is a complicated mix between one of the greatest players in franchise history and a true face of the franchise, and the last remaining contract (along with Matt Garza) that can be justifiably called a holdover from the pre-rebuilding ages. In some sense, he signifies the old guard, an aging veteran from President Doug Melvin’s regime, whose previous crime of PED scandal will now be surpassed by the much worse crime of blocking some imaginary “what could be” of either Corey Ray, Brett Phillips, Lewis Brinson, or Trent Clark (depending on who you ask). Nevermind that there are bountiful universes in which Braun can coexist within an outfield involving any (or all) of those players, as there is no reasonable expectation of linear performance from these prospects over the next four years. Yet there is no middle road in this matter: the Brewers either must hang on to Braun as a solid veteran to lead the next (current!) winning club, or he must be sacrificed to the gods of tanking to return a couple-of-whatevers to bolster some more imaginary “what could be” for that future (whenever!) winner.

Braun makes this state of affairs much more interesting by perfectly splitting both camps:

• His Total Average (TAv) is once again improving, from .278 in 2014 to .298 in 2015, to .316 in 2016, and now .327 in 107 plate appearances thus far in 2017. The aging veteran can hit!

• He finds himself on the disabled list with those famous Braun-brand nagging injuries, this time a neck-and-a-calf.

Given these facts, Brewers fans can dig in their heels and find evidence to support whatever they want to believe on the future of Braun: he remains a fantastic batter which surely means the Brewers must keep him, and he remains injury-prone which surely means the Brewers must trade him.

Last October, I tackled the problem of designing aging expectations for Braun, given the facts that we already know about the slugger.

• Whether one uses the delta model aging curve, harmonic mean, or a multiple regression technique to design an aging curve, Braun is already beyond his peak and (depending on model) has been in decline for three-to-five years. My discussion is not an aging curve.

• Of course, Braun has not declined in a linear fashion, which shouldn’t necessarily be surprising both because Braun is an elite baseball player and (more generally) because individual instantiations of a population can exhibit variances, and so discussing a forward-looking outcome is ever-more difficult.

• There are selection and survivorship biases evident in nearly any cohort one constructs to analyze Braun’s aging development (these are defined in the models I linked above).

• Moreover, Braun simply represents a strange group: his updated and adjusted 2016 WARP of 4.6 places him among the best age-32 players in history; searching Batter Seasons, I found 88 seasons over the last thirty years in which an age-32 player produced WARP of 3.6 or better; if he continues his improved TAv throughout 2017, he will join a group of only 15 (of those original 88) players that posted improved TAv from their age-32 to age-33 seasons.

• Any way one looks at Braun, he’s an outlier, whether it is an outlier beyond the typical aging curve and peak, or an outlier among age-32 and age-33 players, etc. We will almost certainly color our analysis by “knowing what we are looking for” (J.C. Bradbury’s regression methodology discussion features an interesting riff on this theme).

In order to sharpen my previous results, I constructed a batch of Batter Seasons according to a number of principles:

• I limited my search to the last 30 years, which conveniently aligns the late-1980s live ball explosion, the so-called “Steroids Era,” the television revenue explosion of the late-1980s and current decade, and a relatively stable era in terms of medicine, training, etc.

• I looked for players that were age-32 or older.

• I looked for players that produced WARP at 3.6 or greater (3.6 is within approximately five percent of Braun’s originally published WARP that I used in October 2016).

• Among these players, I added in subsequent seasons beyond the 3.6+ WARP performance in order to describe aging performance.

• I then separated these players into two groups: a full group that comprises 935 total seasons, and a smaller group (584 total seasons) that represent players that specifically produced their 3.6+ WARP performance in their age-32 season (rather than age-33, age-34, etc….or age-38 like Brad Ausmus!).

It’s important to take this information with a grain of salt, as I am not using a delta or regression model in order to structure the aging progressions. This is because this group of players already stand beyond the aging curve, and also because they most certainly represent a biased sample that does not necessarily provide conclusions for some broader population of baseball players (for example, if one finds that 3.6+ WARP players at age-32 tend to improve between their age-34 and age-35 seasons, that finding would not necessarily apply to all age-32 players, or even most age-32 players). So, this method is a description of what has happened, rather than a properly structured method that could be used to predict aging progressions. Hence I’d call this a batch or a cohort of baseball players rather than a sample; this is a particular search of very good baseball players.

Table One: All 3.6+ WARP Players After Age-32, 1987-Present

Table Two: Age-32 3.6+ WARP Cohort

There are a few interesting aspects of these tables:

• Players within Table Two predictably drop off in performance faster than the large group because new players are not entering at age-33, age-34, age-38, etc., as they are in Table One.

• Even with this fact in mind, neither group of players ages uniformly or in a linear fashion.

• It is particularly interesting that the players in Table Two run through two cycles of improvement and decline, from 32-33-34 and then 35-36-37. Players from this group that “survive” until age-39 once again improve (slightly) from their age-38 season.

• Ultimately, this is a very good group of baseball players. The median performances from age-32 onward among these players is worth more than $130M based on one free-market pricing of WARP.

• If Braun matched the median of this group during his remaining contract (age-33 to age-36), he would produce approximately $65M of value. The median age-36 performance would be worth approximately $10.9M on a $13M payroll hit ($3M deferred), which would hardly be the albatross that some fans believe.

• It should be emphasized that a Ryan Braun that can produce anywhere near 200-300 plate appearances or 1.0 WARP+  in 2020 can be an extremely helpful asset even on a ballclub that could feature age-27 Brinson or Phillips, age-26 Ray, or age-24 Clark. The development of these prospects will not be linear, their roles are quite unclear, they remain high risk prospects in many cases, and they could easily share time with injuries, ineffectiveness, or platoon roles considered; there is the additional issue that these young players could provide better trade return than Braun, which could greatly help a competitive club address weaknesses and turn itself into a contender.

• It must be noted that even if Braun improves from age-32 to age-33, that does not necessarily indicate anything of significance for his future performance. This would place Braun in the company of Robin Yount (!!!), Adrian Gonzalez, Brett Butler, Chipper Jones, Ichiro, Jason Varitek, John Olerud, Ken Caminiti, Luis Gonzalez, Marco Scutaro, Mike Lowell, Milt Thompson, Rickey Henderson, Roberto Alomar, and Tim Raines, which is a group that has wildly varying success after age-34.

For fun, I tracked the raw change and percentage change of TAv and WARP between age-32 and age-36, pairing each player’s seasons (ex., Robin Yount 32 & Yount 33, Yount 33 and 34, etc.).

Where player seasons were missing, as with active players like Miguel Cabrera or Edwin Encarnacion, or retired players, I added a median “dummy” stat line in order to consistently track changes across 88 players. I added an overall 32-to-36 category to place annual changes in full context.

Here is the shareable Google Drive link for anyone that wants the data:

https://docs.google.com/spreadsheets/d/1y0WB82v1L6q8Ipu2dccLudIkkUxAkkAz2ie4lMQI2TY/edit?usp=sharing