Bannister might have made the difference

In the 2007 season, these pitchers served as the Mets' fifth starter:

Jorge Sosa
Mike Pelfrey
Brian Lawrence

Others served in that role, but the above three were the only ones to win at least one game they started. Sosa won 8 (losing 6), Pelfrey won 3 (losing 9), and Lawrence won 1 (losing 5), for a total of 12 wins. They won 13 of the 33 games they started or 39%.

Coincidentally, in his starts with Kansas City, Brian Bannister won 12 games (losing only 9) for a Royals team that finished the season at 69-93. So Bannister had a won-lost percentage of .571 for a team whose W-L percentage was only .426.

Bannister won won 12 of the 27 games he started or 44%.

If Bannister could achieve a winning percentage on a losing team, imagine how many more games he would have won if he were the Mets' fifth starter. In fact, if he won just two more games, something I believe he was quite capable of doing, the Mets would have ended the season in a tie for the division championship (44% of 33 is 14.5).

Why two more rather than one? The player the Mets traded Bannister for, Ambiorix Burgos, won a game in relief.

Trading Bannister was one of Omar Minaya's biggest blunders last season.