Hey, there’s room for all when Sabermetrics get discussed. Not enough of us around here anyway.
In the interest of knowledge, I went to baseball-reference.com, took the AL 2002 offensive stats for each team. Typed the BA, OBP, SLG, and R/G for each team onto an Excel spreadsheet. Computed each teams OPS and OPS-BA. Correlated each with R/G for the league for that year. OPS won, .967 to .937.
Did the same with the 2002 NL stats. OPS won again, .896 to .810.
Could someone pass the salt? This crow on my plate could use a little seasoning. 
Pitchers on the same team most assuredly do NOT get the same levels of offensive support. Not always, anyway, and the differences can be pretty striking.
Anyway, it’s obviously ERA. The question you have to ask yourself is; how many games would the team have won if they had had an AVERAGE pitcher in there, rather than this guy?
For instance, Roy Halladay is an awesome pitcher, maybe the best in the A.L. unless Pedro is healthy. He’s 20-6. However, I don’t think an average pitcher would have gone 13-13 in those games, because the Blue Jays score a huge number of runs (and, in fact, have scored very well when Halladay has pitched.) If you gave all those starts to an average pitcher, he’d be something like 14-12 or 15-11. So Halladay has really won 5 or 6 games more than an average pitcher would, not 7. That’s Halladay’s marginal value to the Jays - about 5 games, maybe 6.
(An interesting illustration of this is Halladay’s teammate Cory Lidle. Lidle is a way, way below average pitcher; his ERA is 6.02, and that’s no fluke, he’s gotten pounded. But he’s a .500 pitcher, 12-12, because the Jays score enough runs to bail him out.)
On the other hand, take Kevin Brown. Kevin Brown is also a very fine pitcher, and presently he is 13-8. .500 would be 10.5-10.5 - yeah, odd numbers are ugly. But Brown is really much better than 2.5 games above average. The Dodgers are a horrible hitting team, the worst in the National League by a huge margin. An average pitcher would be like 8-13. So Brown is actually 5 games better than average - about the same as Roy Halladay.
Clearly, the fact that Halladay has so much better a W/L record than Brown is entirely due to the fact that he plays for a team that scores a much, much higher number of runs. Halladay has actually added no more wins to his team than Brown has. If you switched them there’d be no doubt the two teams would be roughly even for the trade. There is just no way Halladay is actually nine games better than Kevin Brown.
Of course, to do a truly good analysis I’d have to know how many runs the Blue Jays and Dodgers have scored in those pitcher’s game in particular. You also have to consider defensive support; I’m kind of shortchanging Halladay because the Jays are a horrible fielding team, so actually they are not as good as the offense would suggest. But this illustrates why ERA is more important than W/L.