(Note: Okay, so I changed the name of the post. I'm not always good at coming up with the right name for a post as I'm writing it, and this is a great example. The new title is so much clearer, I think. Hope it doesn't bother anyone.)
I finally splurged a couple of weeks ago and dropped the $25 for the absolutely-worth-it Rally's Historic WAR Database. For those of you who don't know, Rally, the guy behind the CHONE projections system, made available Wins Above Replacement (WAR) data for every player in Major League Baseball history. You can explore the data online at his Player Index or download it all in an easy-to-use database format for a small donation. All of the data is available online, but I like to play with it on my own, so I just had to get the download. Like I said, it's plenty worth it (and it makes me feel a lot better about doing these historical comparisons when I get to use WAR instead of Win Shares).
Anyhow, I've been playing around with the data ever since I downloaded it, but I haven't had much time to write anything about it. I'm finally feeling a little more in control of things, so what better time than now to write about my first study, the Seasonal Age All-Stars?
Using the historic WAR database, I classified every player-season by seasonal age and primary position and then recorded the player's WAR value for that season. "Seasonal age" refers to how old the player was on June 30 of that year, and "primary position" was determined by the position the player played most that season. Once I had that data calculated, it was pretty simple to find the top performers, by WAR, for each age & position combination and, from there, to compile an All-Star team for each age. For example, the Age-19 All-Stars would look like this:
Age: 19 Player Year WAR P Gary Nolan 1967 6.8 P Dwight Gooden 1984 5.4 P Bob Feller 1938 5.3 C Harry Chiti 1952 0.5 1B Jimmie Foxx 1927 0.9 2B Joe Cronin 1926 0.3 3B Buddy Lewis 1936 2.1 SS Travis Jackson 1923 2.3 OF Mel Ott 1928 3.5 OF Ty Cobb 1906 2.6 OF Ken Griffey 1989 2.8 DH Claudell Washington 1974 1.5 Total 34
As you can see, outside of the pitchers, the 19-year-old squad is a little weak. It doesn't stay like that for too long, though, as the Total WAR values increase pretty quickly as the teams get older. The full list of Seasonal Age All-Stars can be found here.
(click "Read More" to keep reading)
A note about the lineup construction: I gave every roster one player at each position, regardless of ties at some positions. I went the generic route for the outfield, just choosing the top three guys instead of trying to break it down by LF-CF-RF (the position data that I have is inconsistent in classifying OFers by position, and I didn't feel like trying to break it down more). I did allow three pitchers per team, and decided to stick with the DH.
The question that inspired this study was "how good would a lineup comprised only of the best age-21-or-under seasons be?" but it quickly grew to more: would the best 19-year-olds be able to beat the best 40-year-olds? what seasonal age would be able to field the best lineup? is it age-27, or might some other age sneak to the top? You can click on over to the spreadsheet to see the specifics, but here is a good summary:
Season Age All-Stars,
Ranked by Total WAR
Rank Age WAR 1 27 123.2 2 24 121.4 3 28 121.1 4 31 118.6 5 32 118.2 6 25 117.8 7 26 115 8 30 114.7 9 33 113.1 10 29 112.9 11 23 108.6 12 34 103.2 13 35 101.5 14 22 100.6 15 21 93.1 16 36 91.9 17 37 87.5 18 20 79.9 19 38 77 20 39 68.1 21 40 56.4 22 19 34
Or, in graphical form:
Seasonal Age All-Stars, WAR vs. Age
It shouldn't come as a surprise to see the age-27 and age-28 teams at the top of the chart - those are the accepted peak years for a ballplayer, after all. I do find the big age-24 peak and the age-26 and age-29 valleys to be interesting, though. And why are the age-31 and age-32 values so high? Could that have something to do with free agency, or is it just a product of the particular players on those teams?
I think that's enough for now. If you haven't already, I highly recommend taking a look at the various Seasonal Age All-Star teams. Some of the names are bound to surprise you (Randy Velarde put up the best ever age-36 season at 2B? Javy Lopez was the best ever age-32 catcher?), and it's always interesting to see just how much Barry Bonds skewed the age-36+ numbers (only two players have WAR values greater than 8.0 after age 36 - Honus Wagner with 8.1 at age 38 and Barry Bonds, with 3 seasons greater than 10.0 between 37 and 40). But I'll leave you with the top two rosters of the Seasonal Age All-Stars and you can tell me which team you'd rather have:
Age: 24 Age: 27 Player Year WAR Player Year WAR P Walter Johnson 1912 11.8 Steve Carlton 1972 12.2 P Christy Mathewson 1905 9.1 Sandy Koufax 1963 10.8 P Hal Newhouser 1945 8.9 Bob Feller 1946 10.1 C Johnny Bench 1972 9.1 Darrell Porter 1979 8.4 1B Lou Gehrig 1927 12 Lou Gehrig 1930 10 2B Rogers Hornsby 1920 9.4 Eddie Collins 1914 11.3 3B Mike Schmidt 1974 10.5 Ron Santo 1967 10.2 SS Alex Rodriguez 2000 11 Ernie Banks 1958 9.7 OF Ty Cobb 1911 11.4 Ted Williams 1946 11.8 OF Tris Speaker 1912 11 Stan Musial 1948 11.5 OF Mickey Mantle 1956 12.9 Carl Yastrzemski 1967 12.2 DH Jim Rice 1977 4.3 Travis Hafner 2004 5 Total 121.4 123.2
So, who would you take?