Musings from a frosty winter day: Tom Glavine, after 16 years and 242 wins with the Braves, has left Atlanta and joined the Mets. How will he fare in New York? Glavine's getting a little old, but great players have a habit of surprising us, and Glavine's longevity has been remarkable to this point.

But the burning question is this: if Glavine struggles, will he move to the bullpen? To my knowledge, Juan Guzman is the only pitcher who never made a relief appearance in his career. Guzman pitched in 240 games, all of them starts. Glavine has a chance to become the first Hall-Of-Famer to never pitch out of the bullpen. Glavine has now made 505 starts, and has never made a relief appearance.

Before the days of six-man bullpens, it was fairly common for starting pitchers to make the occasional relief appearance. Walter Johnson made 136 relief appearances in his career; Bob Feller made 86; Whitey Ford made 60; Bob Gibson made 44; Tom Seaver had only nine appearances out of the bullpen.

As the bullpen has become more specialized, starting pitchers have made fewer and fewer relief appearances. Tom Browning started 300 games in his career, and made only two relief appearances. Roger Clemens made one bullpen appearance in his rookie year in 1984, and that is the only one in his career. So far, Mike Mussina has made 355 starts without a relief appearance.

This is kind of trivial, but these kind of things still interest me. For many years now I have subscribed to Earl Weaver's theory that the best place for a rookie pitcher in is long or middle relief. In that role, a young pitcher can get some experience without throwing too many innings, or pitching in pressure situations. After a year of long relief, a pitcher can then either move into the starting rotation, and perhaps the closer's role. Dennis Martinez, Mike Flanagan, Scott McGregor and Storm Davis all began their careers pitching long relief for Weaver.

I think it is a great idea, and my home team, the Blue Jays, have had a lot of success with it. Jimmy Key, Pat Hentgen, David Wells, Roy Halladay, Todd Stottlemyre, Al Leiter and John Cerutti all spent at least half a season pitching long relief before moving into the starting rotation. Other pitchers around baseball who began their careers in middle relief include Pedro Martinez, Curt Schilling, Chuck Finley, Kenny Rogers, Denny Neagle, Derek Lowe, Jesse Orosco and John Franco.

Of course, lots of pitchers begin their careers in the starting rotation and have success. Glavine, Mussina, Clemens, Greg Maddux, Randy Johnson, Kevin Brown, John Smoltz, Tim Hudson, Barry Zito, Mark Mulder and others began their careers with in the starting rotation. Staying with the Blue Jays, Dave Stieb, Jim Clancy and Juan Guzman were all starters from the beginning, while Billy Koch was a closer in his rookie year.

My thinking is that, while it's nice to get a Mike Mussina or Barry Zito who you can throw right into the starting rotation, for 90% of pitchers it is better to start in middle relief. Most teams, of course, don't really have a policy; young pitchers get fit into whatever roles they are needed. The Blue Jays, I believe, had some sort of policy from 1980-1994; in that period, almost all of their best young pitchers (with the exception of Juan Guzman) began pitching in middle relief.

And that's why Billy Beane and the Athletics irk me because they have done the opposite, and with great success. Like most baseball fans I have tremendous respect for the job that Beane has done, and I can't ignore the fact that, when he gets a good young pitcher, he sticks him right into the starting rotation. First Hudson, then Mulder and Zito, and last year Aaron Harang. The Athletics have a plan, and it works.

This same policy has failed miserably for the Blue Jays since the 1994 season. The Jays in the Gord Ash era threw one young pitcher after another into the frying pan and saw them quickly burn out. And don't ask their respective teams about Kerry Wood, Matt Morris, Bill Pulsipher, Jason Isringhausen, Alan Benes, Gil Meche and Rick Ankiel, all of whom began brilliantly before injuries interrupted or ended their careers. Marlins fans are currently getting the jitters about A.J. Burnett and Josh Beckett.

So, is it just the quality of the pitchers that make the difference? The quality of the organization? Does it make sense to break a young pitcher in slowly before sticking him in the starting rotation? Or are you just wasting a great young arm on junk innings? My guess is that, if nothing else, teams should have a policy, one way or the other. Too often, the focus is on trying to fill a need for the team, and not on doing what is best to develop the pitcher.

It is also interesting to note that starting pitchers have been making fewer relief appearances as bullpens have become more sophisticated. Has the evolution of the bullpen reached its end? Every pen is about the same now: a closer, a couple of setup men, a couple of lefties and a long reliever. Save totals and innings pitched have all leveled off. Is this it?

The development of the bullpen has also pushed down on innings pitched totals for starters, and also complete games. If bullpens are becoming stable, then perhaps these downward trends will level off as well. For the first time in decades, perhaps ever, pitching staffs fifteen years from now may look exactly the same as they do now.