The analyses of pitcher performance relative to his team have yielded some interesting results.

As I described in the "The Theory of Relativity" post, it is possible to compare a pitcher's W-L record and winning percentage to his team's and adjust for factors that distort the comparison. These adjustments involve adjusting the pitcher's run support to equalize it with the run support the team provided to the other pitchers on the team and normalizing the ERAs and runs allowed by the rest of the staff to league average. These two adjustments assure that a pitcher won't benefit or suffer by virtue of run support that deviated from team average, or by virtue of the fact that the rest of the pitching staff, to whom the pitcher is effectively being compared, were either better than league average or worse than league average. A pitcher may be a great pitcher but his W-L record relative to his team's won't be very impressive if the rest of the team's pitching staff is comprised of great pitchers. In comparing Greg Maddux's W-L record to his team's it is obviously necessary to adjust for the fact that the Braves' pitching staffs were great and produced tremendous winning percentages because of the presence of guys like Glavine and Smoltz.

Here are the new results. Curt Schilling's ten-year peak from 1997 to 2006 was pretty impressive, as his excellent winning percentage and ERAs would suggest. Schilling outperformed his team by approximately 27% over that decade. Bob Gibson outperformed his team by approximately 28% over his nine-year peak of '64 to '72; not sure whether people will find that disappointing or impressive. Both these results obviously cast Ron Guidry in a very good light, because Schilling's and Gibson's team relative performance figures are right in Guidry territory. Ron is in good company.

Here are the results I found surprising. I ran the numbers on Bunning, Drysdale and Palmer. I've always thought of Bunning and Drysdale as being very similar, and I've conceived of Palmer as an American League version of Tom Seaver, although a shade behind Tom Terrific. My team relative analyses have fundamentally changed my appraisals of these pitchers.

I didn't expect Drysdale to fare very well in this analysis, simply because his a straight comparison of his W-L records to his teams' isn't very impressive. I expected, however, that Drysdale would benefit from the fact that the Dodger pitching staffs were generally exceptional during his era. Drysdale came in at 9.75%. Now remember that Dave Stieb came in at approximately 17%. That's a big difference. I know Drysdale made in the Hall in large part because of his participation in a lot of great pennant races and World Series, but those who question Drysdale's HOF bona fides have new ammunition for their argument.

Bunning polled in at 16%. That's a pretty good figure, and it's far better than Drysdale.

Now here was a bit of a stunner. A straight comparison of W-L records for Jim Palmer and his Oriole teams never looked all that impressive, because those Orioles teams posted great records during Palmer's peak from '70 to '78. Still, I expected that Palmer would benefit greatly from the fact that those were great pitching staffs the Orioles fielded in the '70s. Well, they were great staffs in the early '70s, but from '74 to '78 they really weren't all that good when you take away Palmer. For the full nine-year period the Orioles posted a 103.6 ERA+ when you subtract Palmer's ERAs. Good, but not great. After adjusting for the quality of the Oriole pitching, Jim Palmer outperformed his team by 20.4%. That's better than Bunning, but not by much. And it's nowhere near Tom Seaver's neighborhood of 37%.

O.K., Jim Palmer wasn't Tom Seaver. Does that make me reconsider whether Palmer was a legitimate first-ballot HOFer? No, not at all. Eight 20-win seasons in nine years is quite an accomplishment, and Palmer was an undoubted big-game pitcher, posting some pretty impressive pennant race performances and superlative post-season numbers. But he wasn't as good as I thought he was.

As with any other statistic or metric, it's important to put it in context. It is an all too common failing of many fans who fancy themselves sabremetricians that they attach too much weight to a single statistic. Schilling comes out well ahead of Palmer in team relative performance. That tends to confirm what Schilling's great winning percentage and ERAs had already told us: Schilling was a damn good pitcher. But Palmer was rock-solid consistent and was generally a bigger winner than Schilling over their respective peaks, even after adjusting for the difference between the four-man rotation that Palmer pitched in and the five-man rotation Schilling pitched in. Palmer had one stinker in '75 when he had significant arm issues, but was excellent every other year. Every pitcher gets a pass for one season where he had arm issues; Seaver had one during his peak, Guidry had one in '84, and most other great pitchers also had one. Schilling had more than his share of them, however, and was significantly limited in his contribution to his team in '99, 2000, '03 and '05 as a result of poor performance or limited innings due to arm issues. He was excellent when he was on, like Saberhagen, but like Saberhagen you can't just ignore all the seasons where his team didn't get what they justifiably expected.

Bunning would appear at first glance to have had more than his share of off seasons during his peak, but that's deceptive. Bunning was very durable and never really missed time due to injury issues during his peak. Some of his pedestrian W-L records, particular in 1960, are easily explained: poor team, terrible run support. There's reason to believe that if you put Bunning on those '70s Orioles teams he might have posted another four or five 20-win seasons. But there's also reason to believe he might not have. It has to be noted that Bunning pitched on some good teams that gave him pretty good run support, like the '61 Tigers and '64 Phillies, and he didn't post 20-win seasons. He did well enough, posting good ERAs and winning 17 games in '61 and 19 in '64. But a Hall of Famer should have been winning 20 games, and probably 22 or 23. Bunning had a bit of a tendency to pitch to his team's level. That's not a particularly damning observation, but it's an important consideration.

Back to Guidry for a moment. The myth is that Guidry's spectacular W-L record was in large part a function of great run support from a powerful Yankee offense. It's a myth. The '77 and '80 Yankee teams could hit, no question, but the Yankees of the pennant winning years of '77 to '81 were unquestionably pitching and defense oriented teams. Look at the stats. And Guidry's run support from the Yankees was in any event strictly average for those teams. Guidry outperformed those teams, and those exceptional pitching staffs, because he was an excellent pitcher, made optimal use of his run support, was a clutch pitcher who pitched best in critical situations, and was in short a winner. That term - winner - is a term a lot of the self-styled stat geeks scoff at, but if they looked a little closer at the stats the concept would be plain. Some pitchers are winners. Some aren't. Blyleven wasn't. Steve Rogers wasn't. To a lesser degree Stieb wasn't either. It doesn't mean they weren't good pitchers. But it means they weren't as good as their generally superior ERAs and ERA+s would suggest. The stat geeks should ponder that for a moment.