July 26, 2000
The Daily Prospectus
Watching the clips of yet another great Pedro Martinez performance in yet another close game got me thinking about the relative stress levels of pitchers. It certainly seems that Martinez has pitched in an inordinate number of close games this season, which means that most of his innings have been fairly high-stress, at least in terms of game situation.
What I didn't know is what a "normal" pitcher goes through, what type of stress level is high or low for your average starting pitcher. Without that information, it's hard to gauge whether Pitcher X is pitching in situations are unusually high- or low-stress. So as I am wont to do, I called upon someone smarter than myself, in this case BP's resident database guru Keith Woolner.
Using 1999 data, Woolner generated a list of pitchers and the innings they pitched, sorting them by the score of the game at the time the pitcher took the mound. If Mike Mussina took the mound with the Orioles up three runs, that inning went into the "+3" column. If Andy Pettitte started the third inning of a 1-1 game, that went into the "0" column. And so on.
Once we had that data, we sorted the pitchers by game situation, labeling one-run and tied situations as "close" and games with differences of four runs or more as "not close." By that metric, here are the pitchers who pitched in the highest-stress situations last season (minimum 100 innings begun):
Pitcher Inn Close Not Close
I confess, I was surprised at how high some of these numbers were. One possible explanation is that pitchers who don't work deep into games--and the first three names on this list qualify--have a higher percentage of their "innings started" in early-game situations, when score differences tend to be smaller.
But even at that, you have pitchers like Jamie Moyer and Mike Mussina starting two-thirds of their innings in close games, or someone like Chan Ho Park pitching almost 14 times as much in these situations as in low-stress situations.
Which pitchers had the fewest high-stress innings?
Pitcher Inn Close Not Close
A mixed bag: some of these guys had lousy seasons, and others had great run support. Because of the "early-game" factor, 40% is probably the lower bound on low-stress innings.
Looking at these numbers, I think you have to say that, oh, Andy Pettitte had a much more stressful season than teammate Hideki Irabu, and that the differences in the situations in which they pitched were significant.
How meaningful is this? I'm loath to draw conclusions based on one season, but I am willing to say that we may need to factor this data into both our ongoing considerations of pitcher workloads and our evaluations of pitcher performance.
Joe Sheehan can be reached at firstname.lastname@example.org.