It seems to me that - for decades - the way a young starting pitcher was brought up was through the major league bullpen. This was a time-honored baseball "unofficial rule" for top prospects. I went to Fangraphs to find out a little more about this. I am going to look at 21 of the 22 men in baseball history to ever throw over 3,000 innings and have a FIP- (FIP adjusted to league and ballpark with 100 being average - the lower the better) of 85 or under. The 22nd pitcher is Andy Pettitte, he is still active. If pitchers can throw 200 innings a year for 15 years and have a FIP 15% better than the league for that long of a time, they are some of the best pitchers ever. I want to know how the best of the best made their transition?

I know there are plenty of examples over the years of pitchers coming up through the bullpen. Just going with Cardinals, because I know them the best, I could find several.

• From 2008-2001, Lance Lynn started 72 of 77 games in the minor leagues. He came up to St. Louis for the first time in June of 2011. He started two games before being relegated to the bullpen for his final 16 regular season appearances and all 10 of his post-season appearances. In 2012, he made the transition to full time starter.
• From 2006-2008, Jaime Garcia started 61 of 62 games in the minor leagues. He came up to St. Louis for the first time in July of 2011. He started only 1 of 10 games in two call ups. In 2010, he was brought to the majors for good; he has started all 71 games in which he has pitched since then.
• From 2000-2005, Adam Wainwright 135 of 138 games in the minor leagues. He came up to St. Louis for a 2 game call up in 2005, relieving both times. In 2006, he was a prime reliever, throwing 61 games and turning into the closer when Jason Isringhausen faltered down the stretch. He finished 10 games in the regular season, earning 3 saves, before finishing all 9 games (saving 4 of them) in which he pitched in the World Series run. Since 2007, he has started every game in which he has pitched and has arguably been a top 10 pitcher in the league and the top pitcher on the St. Louis staff.
• From 1977-1981, Dave LaPoint started 98 of 107 games in the minor leagues with Milwaukee and St. Louis farm clubs. He had call ups in 1980-1982 between the two clubs that saw 50 appearances from LaPoint. He started only 26 of those appearances. Then, over the next three years, he started 93 of his next 101 games.
• From 1971-1974, John Denny started 78 of 81 games in the minors. In 1974, he was called up for a 2 game stint, throwing 2 innings in relief. He then went on to start in all but one of his final 324 appearances in the majors.
• From 1967-1969, Reggie Cleveland started 75 of 76 games in the minor leagues. In 1969, he was called up to start one game. The next season, he threw 16 games for the Redbirds, starting only 1 of them. After that, he made 99 appearances, all as a starter, with the Cardinals before moving on to play for three more teams over the following 8 seasons.
• From 1951-1954, stats are hard to come by - Larry Jackson started at least 90 of his 137 games in the minors, although it seems like it was closer to 120 or more, based on innings pitched. In any case, he was primarily a starter in the minors for four seasons before moving to the major leagues in 1955. In his first four years in the majors, Jackson started only 71 of 178 games played. He also finished 63 of those games - nearly as many as he started. For the final 10 years of Jackson's career, he started 358 of 380 games for 3 different clubs.

Needless to say, there is plenty of precedent for this "unwritten rule" to have had effect.

One player that I left out of my list of Cardinals above is Bob Gibson, the greatest Cardinal hurler of all-time. Gibson started only 21 of his first 40 games as a major leaguer. Let me re-iterate: The greatest Cardinal pitcher of all-time started only 52.5% of his appearances in the first TWO years he was in the majors! This is a guy who threw 17 seasons for one team. He threw 3,884 1/3 innings of 78 ERA- and 84 FIP- ball. He struck out nearly 20% of all the batters he faced in 17 years, while walking less than 8.5% of them.

Back to the top 21, though. The 21 players that I referenced earlier are (in order of FIP-):

1. Roger Clemens*
2. Randy Johnson*
3. Curt Schilling
4. Walter Johnson*
5. Christy Mathewson
6. Lefty Grove
7. Greg Maddux*
8. John Smoltz*
9. Kevin Brown*
10. Pete Alexander
11. Bert Blyleven*
12. Bob Gibson
13. Mike Mussina*
14. Cy Young*
15. Mordecai Brown*
16. Chief Bender*
17. Dennis Eckersley (how many of you knew he came up as a starter?)
18. Rick Reuschel* (I had no clue watching him growing up that he was this good!)
19. Nolan Ryan*
20. Paul Derringer
21. Tom Seaver*

Those 21 pitchers averaged only 19 starts in their first 29 games. That is a pretty good stat there. Out of 21 of the best pitchers of all time, they averaged starting only 65% (approximately) of their early on in their careers. However, the ones with an asterisk (*) above started at least 80% of their games - most over 95%. Out of the remaining 7 (Schilling, Mathewson, Grove, Alexander, Gibson, Eckersley, and Derringer), they only averaged starting only 21 of their first 50 appearances (only approximately 40% of their starts). So 1/3 of the top 21 pitchers listed here started merely 2 of every 5 games they played early in their careers.

These players' careers spanned from 1890 until 1991 at the front end and 1911-2009 at the back end. So for a century, this was quite an accepted trait in the majors and worked out incredibly well for many, many pitchers in their careers. What I am surprised at in today's game is how little it seems to happen. It seems to me that the fluidity of bullpens in the past 5-10 years has grown exponentially in this "what have you done for me lately" world of sports in which we live. There are 7 spots in the bullpen on each team each season. That means there are 210 spots in major league bullpens at any given time of the season. Even if one player from each bullpen were to retire each year, which would mean teams should have used 16 pitchers in 10 seasons. If you assume injuries each season, but overlap of pitchers year-to-year, it might be reasonable to assume that teams should use 25 pitchers in their bullpen in 10 years. The average major league bullpen from 2003-2012 (10 seasons) has used 54.4 pitchers. In the 10 seasons prior (1993-2002), only 48.6 pitchers were used per bullpen. In the 10 seasons before that (1982-1992) the number was under 40 bullpen arms. With such fluidity, why wouldn't this strategy be employed MUCH more often right now - in today's game?

There have been 18 pitchers (not including Randy Johnson and Mike Mussina, used above) in the last 10 seasons to have thrown at least 1,000 innings of 85 or less FIP-. They are:

1. Tim Lincecum
2. Roy Halladay
3. Chris Carpenter
4. Adam Wainwright
5. Felix Hernandez
6. Johan Santana
7. CC Sabathia
8. Zack Greinke
9. Ben Sheets
10. Justin Verlander
11. Jake Peavy
12. Roy Oswalt
13. Brandon Webb
14. Cliff Lee
15. Dan Haren
16. Josh Beckett
17. Andy Pettitte
18. Jon Lester

These 18 pitchers averaged starting 22 out of 31 games pitched (71% of their games). Again, 2/3 of the pitchers (with a *) averaged 80% starts or higher (most at 100% or 1-2 starts away from it). Out of the 1/3 that did NOT make their debut starting over 80% of their games, those remaining 6 pitchers started only 20 out of 45 games on average. That's only 44% of games pitched as starts. However, BOTH of these numbers are actually HIGHER than history would dictate.

I'm sure I could do another entire post on the reasons that this is counterintuitive to what I was thinking, but I do not know where to start on that right now or how to go about doing it - nor do I want to make this any longer. There is only one piece I would like to add to this, then.

The pitchers in the minors for the St. Louis Cardinals and Kansas City Royals (at least) who are so highly touted are nearly all starting pitchers. For the Cardinals: Shelby Miller, Carlos Martinez, Tyrell Jenkins, and Trevor Rosenthal are the four highest rated pitching prospects pitching this season primarily in the minors. For the Royals: Jake Odorizzi, Mike Montgomery, Yordano Ventura, and Chris Dwyer are the four highest rated pitching prospects pitching this season primarily in the minors. (I know that Eduardo Sanchez for STL and Kelvin Herrera were top 5 pitching prospects before the season and are both relievers, but they have spent their time primarily (or all) in the majors this year.) The others listed are ALL starting ptichers. Rosenthal is the only one of the 8 to make it to the majors, and he was a relief pitcher for just over a week, before being sent down to join the rotation in AAA. For a team like the Royals, who are out of contention, or a team like the Cardinals, who have such bullpen problems this season (in terms of effectiveness - not "stuff" or "peripherals", but actual outcomes of getting outs before runs are scored) - besides the service time factor, why not bring these players along through the bullpen?

Just my thoughts. Feel free to disagree in the comments section below.