[Cryomaniac]

Originally Posted by Curtis

LGO will probably be by shortly to straighten us out on this, but as a very rough rule of thumb, that's true for the 1870s. By the time the American Association is formed you should probably add a starter, and by the time it folds one more.

A really great book on this topic is "The Great Encyclopedia of 19th Century Major League Baseball" by David Nemec. It has more facts, figures and statistics than articles. I'll just list the number of starters for teams in 1882 and 1892, Both are in order of team's finish.
1882 (you may feel this supports your position):
Chicago (2) 45 and 39
Providence (2) 52 and 32 starts
Boston (3) 48, 32 and 5
Buffalo (4) 51, 29, 3 and 1
Cleveland (3) 67, 16 and 1 (67 led the NL)
Detroit (3) 45, 39 and 2
Troy (3) 42, 33 and 10
Worcester (6) 46, 18, 14, 3, 2 and 1 (26 for 8 teams)

Cincinnati (3) 54, 25 and 1
Philadelphia (7) 41, 20, 8, 2, 2, 1 and 1
Louisville (5) 55, 11, 10, 2 and 1 (55 led the AA)
Pittsburgh (6) 38, 23, 14, 2, 1 and 1
St. Louis (9) 45, 14, 8, 4, 3, 2, 1, 1 and 1
Baltimore (7) 39, 13, 13, 3, 3, 2 and 1 (37 for 6 teams)
1892:
Boston (6) 51, 48, 35, 16, 1 and 1
Cleveland (7) 49, 42, 28, 26, 5, 1 and 1
Brooklyn (6) 44, 42, 23, 21, 20 and 8
Philadelphia (7) 49, 38, 36, 18, 9, 3 and 2
Cincinnati (12) 49, 34, 27, 16, 10, 7, 3, 3, 3, 1, 1 and 1
Pittsburgh (10) 53, 36, 26, 13, 12, 4, 3, 3, 1 and 1
Chicago (6) 71, 45, 26, 2, 1 and 1 (71 led the league)
New York (4) 62, 47, 43 and 1
Louisville (9) 40, 31, 24, 18, 16, 15, 4, 4 and 2
Washington (11) 52, 23, 21, 15, 14, 7, 7, 4, 4, 3 and 3
St. Louis (9) 45, 32, 20, 13, 12, 11, 10, 10 and 2
Baltimore (13) 47, 46, 21, 13, 8, 6, 3, 2, 2, 1, 1, 1 and 1 (100 for 12 teams)

Based on that very small sample, it looks to me like the American Association led the way to larger staffs. Also note that in 1892 the teams are playing close to 150 games, whilst in 1882 the NL clubs were scheduled for 84 games, and the AA clubs played 73-80. As the seasons lengthened, the number of pitchers needed did as well. Someone with more mathematical savvy than I have could determine which percentage of games started by by the number one starter, number two, etc.

Interesting numbers, and yes, I'm sure LGO will have something to say about them. I'm not sure the number of starters is the problem here though, it's more the number of relievers. Again, I could be wrong here, but most of the teams you mention probably only had 1 or 2 actual relief pitchers (sometimes another starter would finish a non-shutout blowout for example). What OOTP should model is teams going from 1 SP + 3RP in 1870 to 5/6 SP + 7/8 RP in 2008, that really doesn't happen, because even if you set up 15 man rosters and 1 man rotations you get 6 pitchers and 9 hitters, usually one of each position plus a backup C.