Everyone has a way to evaluate a player. Give x points for this, and y points for that, convert that into wins, and all of a sudden Mike Mussina is worth 3.84 wins and Clemens is worth 2.98. And based on this, we're supposed to declare Mussina to be the better pitcher. Is that the case though? After all, what about Clemens' incredible winning record? Is that too team dependent?
You have to remember that the methodology people apply is based only on those events that they choose to include. A pitcher's "ability" to pitch in tough situations, or pitching to the score, is almost always ignored. In fact, we don't even know if this ability exists. But if a pitcher gets many 2-1 wins, we'll probably give the pitcher the benefit of the doubt, realizing that the game was on the line on every pitch. And if a pitcher gets an 8-4 win, we might give him a tiny benefit as well, realizing that he might have been pacing himself.
The games
So, let's look at the individual games that these pitchers pitched in. Combined, these pitchers pitched in 67 games. I split their games into 3 more or less even categories: good support, normal support, poor support. Support is defined as offensive run support, fielding support, and opponent "support" (meaning if the opposition was poor offensively, then these teams would be supporting the pitcher). We end up with 21 games with poor pitching support, 22 games with good pitching support, and the 24 normal games.
Mike Mussina was the benefit of 9 games with good support, while Clemens has 13 such games. As it turns out, the Yankees won all 21 games. Mussina had 4 good starts, 3 normal, and 2 poor ones. Clemens had 6 good starts, 3 ordinary one, and 4 poor ones. We could accept that both pitchers are a wash here. These are their numbers when the pitching was not needed:
Pitcher GS IP H R ER HR BB SO ERA Clemens 13 87 88 40 38 12 27 71 3.92 Mussina 9 63 51 22 20 10 10 59 2.86
Poor support
Mike Mussina had the unenviable position of pitching in 15 games where there was poor pitching support, while Clemens only had only 6. In these 15 games, the Yanks scored 26 runs for Mussina, while scoring 10 in 6 for Clemens. The Yanks won 2, lost 3, and tied 1 for Clemens. With Mussina on the mound, the Yanks won 3 and lost 12. Quite a difference. How did these two pitchers pitch when the game was on the line with every pitch? Mussina had 7 games where he pitched good to great, 4 ok, and 4 poor. Clemens had 5 good to great games, and 1 poor. Clemens easily outclasses him here. These are their numbers when the pitching was most important:
Pitcher GS IP H R ER HR BB SO ERA Clemens 6 40 37 15 14 1 17 52 3.18 Mussina 15 90 101 54 49 5 19 90 4.90Mussina gave up alot more runs than would have been expected given his other numbers. Either he had the bad luck of giving up many back-to-back hits, or he had the bad ability to breaking down with men on base and not having any support around him.
Neutral games
Finally, the neutral games, and this is where Mussina shines. 10 starts for Mussina, and the Yanks win 8 of them. Clemens had 14 and they win 12. Clemens had 7 good starts, 5 ok, and 2 poor. Mussina had 8 good to great games and 2 poor ones. These are their numbers when the pitching was important some of the time:
Pitcher GS IP H R ER HR BB SO ERA Clemens 14 93 80 39 34 6 28 90 3.28 Mussina 10 76 50 11 11 5 13 65 1.31Mussina gave up alot fewer runs than would have been expected given his other numbers. The complete opposite of the previous situation. We'd have to give an edge to Mussina in this case.
And the answer is...
If we only look at the last 2 categories where pitching mattered at all, Clemens had 12 good starts and 3 poor ones (and 5 ok). Mussina had 15 good and 6 poor (and 4 ok). Overall, these two pitchers certainly pitched at very comparable levels. Stat-heads love to say how Clemens wasn't at the same level as Mussina, and sportwriters love to say how Clemens was in a class by himself. The answer lies in the middle. I don't see how you can make much of a distinction between the two pitchers.