Designated HitterJuly 30, 2010
The WAR Against Age
By Doug Baumstein

In this article I examine at what ages baseball’s very best hitters had their best seasons as measured by wins above replacement (“WAR”). I looked at the top 40 position players in career WAR and plotted their top 5 seasons against their age during that season. Thus, with 200 data points in all, I created the below chart plotting a player’s personal top 5 season against his age. 

Screen%20shot%202010-07-29%20at%209.14.29%20PM.png
Among superstar position players, personal top 5 seasons occurred anywhere between ages 20 and 39. The median age of a top 5 season is 28, although among the players selected, a top 5 season was generally as likely to occur during the age 26, 27, 28 or 31 seasons. Presumably, the relative lack of age 29 and 30 seasons among the 200 seasons looked at has more to do with small sample size, but the decline in the chart is worth noting nonetheless.

Obviously, looking just at the numbers is not that enlightening, so I also noted some of the more interesting results as they pertain to individual players. For example, the three 20-year old seasons that were among the personal top 5’s of the players on the list belonged to Mel Ott, Al Kaline, and Alex Rodriguez. I, for one, would not have guessed that one of A-Rod’s best seasons was his first complete season. The four 21-year old seasons that make the list belong to Rickey Henderson, Eddie Matthews, Jimmie Foxx and Ken Griffey Jr.

On the other end of the spectrum, the 39 year-old season belongs to Barry Bonds, who likely found his fountain of youth in a syringe. The two age 38 seasons on the list almost certainly had nothing to do with chemical enhancement, as they belong to Cap Anson and Ted Williams. Anson, as it turns out, was not only a great old-time player (even if less than a great human being), but was one of the greatest old players, turning in his best five seasons at 29, 34, 36, 37 and 38. Williams, for his part, of the top 40 hitters, had the biggest age gap among his top 5 seasons, turning in one at 38 and one at 23. That, however, is likely more a function of geopolitics than playing ability, as Williams turned in an 11.0 and 11.3 WAR season at the age of 23 and 22 in 1942 and 1941 and an 11.8 and 10.3 WAR season during his age 27 and 28 seasons in 1946 and 1947. Although, had he played and not fought in World War II there is no guarantee he would have exceeded the 9.9 WAR season he had at 38 years-old (10 WAR seasons are few and far between), had he had one such season during the three years between 1943 and 1945, the 15 year difference among his top 5 seasons would not have existed. Other top players who turned in at least two of their top 5 seasons more than ten years apart include Barry Bonds (age 28 to 39), Tris Speaker (24-35), Al Kaline (20-32), Carl Yastrzemski (23-33), Joe DiMaggio (22-33), Rickey Henderson (21-31), A-Rod (20-31), Eddie Matthews (21-31) and Chipper Jones (24-36).

A number of players put together their 5 best seasons in a row, showing a true peak and incredible consistency. Those players include Hank Aaron (25-29) (I always thought of him as someone who had his best years late, but he actually peaked on the young side), Honus Wagner (31-35) (a renowned older superstar), Joe Morgan (28-32), Wade Boggs (27-31) (I would not have guessed that he was a top 40 WAR hitter, and he was actually number 27, ahead of George Brett (number 30) who I consider the better player), Charlie Gehringer (30-34) and Rod Carew (27-31). A few others put up their best 5 seasons in a six year span, including Roger Connor (27-32), Roberto Clemente (31-36) and Jeff Bagwell (26-31). I find Clemente’s late surge especially interesting. I have always believed that, to the casual fan, Clemente was one of the most overrated players ever. He died after his age 37 season, shortly after the most productive stretch of his career, possibly increasing the halo effect surrounding his untimely and tragic death, and potentially creating a stronger impression of his playing abilities than might otherwise have been deserved had he gone through a typical decline phase.

I also looked at some players outside of the top 40 to see if there were any interesting patterns. Craig Biggio showed a consistent peak, turning in his top 6 WAR years from 28-33. Jim Edmonds showed a late peak, turning in his top five years from 31-35. Paul Molitor’s top five seasons also showed a late peak at the ages of 34, 35 and 36, although his age 25 and 30 seasons also constitute his top 5.

Conclusion

When I started this exercise (and I did look at a lot more stars from the so-called steroid era, even if they were not in the top 40), I expected to see that modern stars, as a result of advances in training, exercise, medicine and performance enhancing drugs would turn in the best “old” seasons. Other than Barry Bonds’ anomalous age 39 season, the evidence seemed to point the other way, as players such as Honus Wagner, Cap Anson and Roberto Clemente all showed later peaks than typical current-day star players. Also, I was surprised that 7 (or roughly one in six) of the superstars who I looked at turned in one of their top 5 seasons at age 20 or 21. While it is not surprising that superstars break in early, it is surprising that many had among their best seasons before they legally could buy a beer in today’s world (although Jimmy Foxx didn’t seem to have a problem in procuring a beer in his time).

I also performed a quick review of post-World War II pitchers. Although I did not find anything all that surprising, pitchers seemed to show a far greater dispersal in value at different ages. Time permitting, I will take a look at that data and prepare a similar study, and see if pitchers age differently than hitters or whether their peak seasons generally occur during their late 20’s.

Doug Baumstein is an attorney and Mets fan living in New York.

Comments

That seedy remark about Barry Bonds was gratuitous.

Hi, thanks for the article. I like your approach, but perhaps expanding from 40 players to a larger number would help smooth out that curve a bit? Regression can really skew things when a guy has a great season followed by a poor one. Also, how did you deal with injury years?

I've done something pretty similar, but instead of sticking with a player set, I took the top 25 WAR at each age, from 2000-09.
http://goo.gl/EmRC

Great work as always. How about this for an early peak: Dizzy Dean. Dean's five best seasons in terms of WAR were consecutive, and went from age 22-26. Dean compiled a WAR of 31.2 for those five seasons. His career WAR was 39.6. Dean's 6th best season was at age 27 w/a 4.5, done in only 197IP. That was the year he messed up his arm while trying to pitch with the broken toe via Earl Averill. So 35.7 of Dean's career 39.6 WAR was compiled in just 6 seasons.

Aaron, thanks for the comment. Although I will address it in my next article, take a look at Frank Tanana, who had tremendous years in his 20-23 seasons.