Clint,

On re-reading my post, I realize that it sounded rather like an ad hominem attack, which was inappropriate. I apologize. I spoke as if you were attempting to mislead, where in fact all I really wanted to say was that in Clarke's case the standard Jamesian ways of looking at peak value are quite unusually ineffective in revealing his peak _ability_. If you compare your avg. value in top 5 consecutive seasons with my 5-consecutive season peak rate, you'll see that the position of the other 8 players on the two lists is quite similar, while Clarke's is very different. His injuries in what would otherwise have probably been the best seasons of a career otherwise exceptionally long and generally durable give him an odd profile, in which his best seasons are widely separated and fall outside his period of peak ability.

Because of these injuries, Clarke's best seven seasons are somewhat lower than the best seven seasons of Flick, Duffy, and Kelley, however you divide them up, as you showed.

However, and this is perhaps the most important point, the fact that Flick, Duffy, and Kelley were slightly better than Clarke in their top seven seasons is outweighed by the fact that Clarke has another 8 above-average seasons beyond those seven, where F, D, K have only 2 or 3 each. If you even go out to their top ten seasons, Clarke passes Duffy and Kelley in total win shares, and he trails Flick by only one. If one is interested in value in a pennants-added sense, there's no way that F,D, and K's slightly higher value in peak seasons is outweighed by Clarke's huge advantage in above average seasons. If we had pennants-added numbers for Clarke, I am confident that he would lead all eligible outfielders by a substantial margin.

Apropos of the Chris/Clint discussion, I posted my top 25 peaks in #102 and noted that Clarke and Keeler are not among them. In #122 I also noted that OF have a hard time competing on peak. Because among the caliber of C-2B-SS-3B whom we are considering for HoM, many at one time or another hit like an OF and also had more defensive value than an OF can accumulate.

So I balance my (3 and 5 year) peak ratings by a prime rating, a period of time that flexes to the player's yearly performance. Here are the top 25 "active" primes including the number of years (numbers of years can vary from WARP to WS). I have STOPPED using rate stats because of their ambiguity as it relates to real impact.

1. Cy Young--15 to 19 years!

Bobby, Bobby, Bobby, read post #102. All will be revealed.

Alright, I've got the top three set on my 1917 ballot:

1. Cy Young: The NL Years (1890-1900)

I know I'll be ducking some rotten tomatoes in this crowd for saying this, but I think if used properly, GPA is a better measurement of hitting value than OPS+. Of course, you would have to adjust all players to a fixed hitting context, but, for me, it more accurately reflects the importance of on-base percentage than does OPS+.

Take a look at the following two players:

Player A: .353 OBP/.519 SLG/.872 OPS

In the Topsy Hartsel comment in the NBJHBA, James creates a rating for leadoff hitters: "... take the number of times the player has been on first base, multiply by .35, his times on second by .55, his times on third by .8 and his home runs by 1". Then convert into Expected Runs Scored per 27 outs, and adjust for league context. The top players on this measure are Williams, Ruth, Mantle, Barry Bonds - not leadoff hitters. Among those who actually were leadoff hitters:

1. 167 Rickey Henderson

I see Clarke, Kelley, Keeler, and Flick as pretty clear HOMers, which frankly makes it difficult for me to care about any other OFs. I'd consider Pike as a last 1870s guy, but he's not a must and I doubt we'll ever run short of OFs again.

Craig B #88

Chris Cobb #126

Another example for Craig and Paul, although not quite as spectacular: Jose Cruz. After a promising start, he had OPS+ of 80 and 91 at age 24-25 for the Cardinals, then a year at age 26 in which he mostly didn't have a job. After being traded to the Astros, he had a 6-year peak at ages 27-32 with OPS+ averaging 125 or so, then he started to fade, then he had two years at ages 35-36 at above 140. At age 36 he hit 13 triples and went 22-8 as a base stealer.

Clemente and Cruz hung onto their gains beyond 5 years and remained strong players into their late 30's. Jennings became a first baseman at 30, a weaker hitter at 31, and a part-time player at 32.

Clarke was a sensation in 1995, his first full season. See his record in context of the awful Louisville team and it is easy to imagine why.

Clarke was a sensation in 1995, his first full season. See his record in context of the awful Louisville team and it is easy to imagine why. http://www.baseball-reference.com/teams/LOU/1895.shtml</i>

25-year-old rookie Jimmy Collins was second on the team with 10 Win Shares in 96 games. Collins made his professional debut 16 May 1893 for Buffalo(EL), age 23y-4m, fresh from the Buffalo sandlots. --Joe Overfield, in Baseball's First Stars (SABR 1996).