We're hearing it everywhere we turn: Pettitte is a big-game pitcher. Pettitte is experienced. Pettitte has proven he can get it done.
In just one new example, msnbc's Mike Celizic writes that Joe Torre did the Indians "a favor by starting Chien-Ming Wang, his biggest winner, instead of Andy Pettitte, his best big-game pitcher."
Turns out that, statistically, Roger Clemens has certainly had the better postseason career in roughly the same amount of work (which we would expect based on career averages). But don't tell that to the scribes!
I find it handy to keep the following stats around and use them to destroy the mindless hordes who spout this "wisdom."
Pettitte career ERA: 3.83
Pettitte postseason ERA: 4.08
Clemens postseason ERA: 3.66
Pettitte career WHIP: 1.35
Pettitte postseason WHIP: 1.35
Clemens postseason WHIP: 1.21
Pettitte career K/9: 6.6
Pettitte postseason K/9: 5.7
Clemens postseason K/9: 7.9
Pettitte career K/BB: 2.3
Pettitte postseason K/BB: 2.3
Clemens postseason K/BB: 2.5
Pettitte career BABIP: .315
Pettitte postseason BABIP: .305
Pettitte career FIP: 3.81
Pettitte postseason FIP: 4.27
Clemens postseason FIP: 3.64
What we find is that Pettitte has performed in the postseason just as we might expect given his career averages. By most measures, he's been slightly worse in the postseason. He most definitely has not stood out as an appreciably better pitcher.
As Jay has aptly pointed out, Pettitte has had nine horrid starts out of 34 total postseason outings. Yankee faithful are more likely to cite his handful of dominant starts; it's easy to remember the things we like.
The second half of Pettitte's 2007 season (the best indicator of what we can expect in the playoffs) has essentially mirrorred his career averages: 3.85 ERA, 1.42 WHIP.
Trying to predict the postseason is folly, but Pettitte's past performance gives no one any reason to believe he'll dominate tonight.