In a recent interview, Dan Duqette, General Manager of the Red Sox from 1996-2002, with reference to that "other stuff" that I'm sick of reading about and don't really want to write about, was asked again about his famous "twilight of his career" comments during his negotations with the Clemens camp, during which he made a non-infamous public comment about Roger Clemens being in the "twilight of his career." The aftermath of this, of course, was Clemens going to Toronto and winning two Cy Youngs in a row in 1997 and 1998. Then he retired on top. No, wait, I mean he pitched 9 more seasons after that and won two more Cy Youngs while appearing in in several World Series, winning two with the Yankees in 1999 and 2000.

In an ensuing discussion at Baseball Primer Think Factory, some commentators wondered whether or not it was reasonable for Duquette to have seen Clemens as being on the decline. In an obvious sense, the answer is "yes," as almost all baseball players (with very, very few exceptions outside of Barry Bonds and Honus Wagner) are in the "twilight of their career' at 33. Again, I'm not going to discuss the "other stuff" that may or may nor have helped Clemens in his post-1996 career. Nor am I really interested in evaluating the state of sabermetric analysis inside and and outside of baseball at that point. What I simply want to do is to see what "twilight" meant at that point. Again, this isn't meant to say what Duquette "should have known," especially since Voros McCracken's discoveries were a long way off. This is just to gain a big of perspective on Clemen's status in the game using more contemporary methods of pitcher valuation such as those used here.

Using FIP (Fielding Indepedent Pitching) -ERA to calculate pitcher WAR (Wins Above Replacement), here are my calculations for the top 5 pitchers in the American League in 1996.

For good measure, let me (without introduction) do the same calculation using BaseRuns-FIP as developed by Colin Wyers, a very cool dynamic FIP that I want to start using more and more and my "go to" pitching stat, and hope discuss in more detail down the road. I've simply droppped it in in place of FIP while using the same methods for getting win%, the same park adjustments, etc. as above. Once again: the 5 most valauble AL pitchers of 1996:

Clearly, the Cy Young voters that season preferred BsR-FIP over the other kind!

There are a lot of factors that go into a decision to resign a player. We haven't gone into the negotiations or Clemens' recent performance. This isn't a career retrospective, or an evaluation of Duquette's decisionmaking. But looking at the numbers above, I think it is fair to say that "twilight of his career" may have been pushing it.