One of the problems of looking at this stuff from the pitching side is that the stats are harder to come by. While on-base percentage and slugging average allowed are available in some of the more esoteric stat books, that kind of data isn't readily available. One source of data is STATS Player Profiles, which gives OBP and SLG allowed by individual pitchers, each team, each league and all of major league baseball combined. These numbers are excellent for getting a handle on just what constitutes "average."As a starting point, here are the OPS allowed and ERA numbers from the American League in 1999.

Team	OPS Allowed	ERA
Seattle	.822	5.25
Kansas City	.819	5.35
Tampa Bay	.817	5.06
Minnesota	.805	5.03
Texas	.805	5.07
Chicago	.800	4.92
Detroit	.800	5.22
Toronto	.800	4.93
Baltimore	.787	4.77
Cleveland	.784	4.91
Anaheim	.773	4.79
Oakland	.773	4.76
New York	.730	4.16
Boston	.713	4.00
Average	.788	4.87

A cursory look down the list shows that -- no surprise -- the teams with the high OPS allowed tended to cough up a lot more runs. A simple table doesn't exactly show the relationship clearly though, so here's what it looks like plotted out:The correlation is actually quite stunning. The only teams that can be remotely described as anomalies are Tampa Bay, who managed to prevent a few more runs than they should, and Detroit, who balanced the Rays nicely by allowing a few more runs than they should have. All the other teams are in a tight little band around the red line.

One of the more interesting features of the chart is the fact that the Yankees and Red Sox are so far away from the competition. This doesn't surprise me for the Yankees, as a fair amount is due to the difficulty of scoring at Yankee Stadium. The Red Sox are a different matter, though.

The obvious factor for them is the presence of Pedro Martinez. While Ken Griffey Jr. might get 10% of his team's plate appearances, A starter like Martinez will get more like 17% of his team's innings pitched. Thus he has almost twice the leverage to reduce his team's OPS allowed and hence ERA. And we all know that reducing runs allowed is just as effective in helping your team win as scoring more runs. (We do all know that, don't we?)If the average ERA is 5.00 (which is just about the case in the AL), Martinez' ERA is 2.00, he pitches 17% of his team's innings, and the rest of the team is average, then Pedro reduces team ERA by 0.31, all by himself. Not too shabby when you think about it. If you take out the "Pedro effect", you'd see that the Red Sox ERA would still be 4.30, well below average still.

In effect, these numbers give us some more ammunition in the argument over the relative value of a superstar slugger and a superstar pitcher. Even though Pedro Martinez might only start once every five days, he still has almost twice the ability to influence his team's ERA. I don't understand why so many people (including many prominent baseball commentators) don't understand the value of pitching (at least relative to hitting) when all of baseball history tells us that it's the pitching, stupid!Oh well. Looking forward to the coming season, it's pretty evident that the Red Sox and Yankees will still be the pitching staffs to beat. The Mariners should at least move from the top spot, as their staff got better and the new stadium will suppress scoring, at least on cold days (of which there'll be many before July, believe me.)One final thought on pitching and OPS allowed. For a single team season, the sample of data is large enough for any serious good luck or bad luck to even out, hence the solid correlation. The same analysis can be done for single players, but the sample size becomes rather small, so the fluctuations, or good and bad luck, as we tend to refer to them, are larger. As with many things, however, good and bad luck tends to even out over the years. So look for a multi-year individual pitching analysis sometime soon.

about the author

Dave Paisley is an actual matriculated engineer, unlike those other stat guys. He even owns a slide rule, so don't even question. Okay, go ahead and question at drdjp@strikethree.com, but make sure to use terms like "Pythagorean" and "regression."

Custom Search