1.  i think i'd have given it to stewart, probably.

2.  i've never calculated RSAA before, but i thought i'd give it a shot. so this might be wrong, but i've got stewart at 45 RSAA, clemens at 51.3 RSAA, and welch at 25 RSAA. maybe someone else can double-check if i've got the formula right. clemens has a good cy young case right there. but it's close, and stewart did have 11 complete games to roger's 7, and stewart did pitch almost 40 more innings and had one more win than clemens. i'm still sticking with stewart, but it's very close. and i don't feel at all bad for clemens because 1)he seems like a jackass and 2)he's gotten some less-than-fully-deserved cy youngs himself.

27 wins is nice and all, though. but i don't like that the cy young is about wins wins wins. for example, i thought randy johnson should have gotten it last year. it's not his fault the diamondbacks were terrible.

3.  I don't really care about wins if it's a difference between 21 and 19 wins. But 27 wins is the most ever in the AL in the DH era. Only two other pitchers have won 25. It's a special achievement that deserves recognition, and the best way to recognize it is with the Cy Young award.

4.  I thank you for reminding me of Stewart's routine Clemens-stomping (sweet, sweet memories!), but I've got to dissent from some of your argument.

I'd agree that "at some point, you gotta reward what people actually do (win actual games), instead of what they might have done". But it's a conceptual and grammatical illusion to say that Bob Welch "actually" won all those games. The wins were attributed to him, by rule and by convention. We all know that the connection between subject and verb in the sentence "Player A won the game" is a lot looser than in the sentence "Player A hit a home run."

I remember, at the time, going over the season's starts, comparing Welch's and Stewart's starts. Stewart pitched much better, winning a lot of tight, low-scoring games (a couple 1-0 games in there), while Welch pitched quite well but above all got boatloads of runs. When a pitcher is on the right side of a 1-0 score the statement that he "won" the game makes a lot of sense; if the score is 9-4 then really the team won the game and the pitcher was the beneficiary.

People knew this at the time, and commented on it regularly. Not for one day was Stewart not considered the ace of the staff, and no one even considered that Welch should lead the playoff rotations. People were aware that Stewart was not only a better pitcher, but was pitching better that very season, and yet most people still felt compelled to give the Cy to Welch because... well, you just had to, if a guy won 27 games. I don't remember my own thoughts on the Cy Young Award at the time; I may well have agreed that you just sort of had to give the thing to Welch.

I sound very negative here, but none of this should construed as meaning I was anything less than thrilled with Welch's season. I loved it.

I didn't love what happened next, which isn't really pertinent to the Cy discussion but was very important to the franchise. Welch was a free agent after that season, and wanted a four-year contract. Sandy Alderson dug in his heels for a long time, offering just three years, but finally caved in with a 4th guaranteed year. I know it was very much against his better judgment, but it's awfully hard to cut loose your 27-6 Cy Young winner. I know I was uneasy about the contract, but relieved to have Welch back on board. With his spiffy new contract, Welch proceeded to give the A's not four, not three, not two, not even one, but zero good years. He went 35-37, giving the team just 579 innings of 4.80 ball. Four Joe Blow years at a star pitcher's salary was a major drain on the franchise.

5.  Let's see.

Welch had 3 starts where he gave up 5 ER or more. He was assigned 0 wins and 3 losses. Clemens had only 1 such games, and Sox lost.

Welch was in 6 games where he gave up 4 ER (total ERA of 5.84), and A's won 4 and lost 2. Clemens only had ONE such games and the Sox won.

3 ER? 5 for Welch (ERA of 3.79) and A's were 4-0, with 1 ND. Clemens had 4 (4.63) and Sox were 2-1, with 1 ND.

2 ER? 6 for Welch (2.55), 5-1. NINE for Clemens (2.53), 6-2, 1 ND.

1 ER? 9 for Welch (1.31), 8-0, 1 ND. Clemens also 9 (1.15), 7-2.

0 ER: 6 Welch (48 innings), 6-0. 7 Clemens (58 innings), 5-0, 2 ND.

At the 0,1,2 ER:
Welch: 21 G, 1.24 ERA, 19-1, 1 ND.
Clemens: 25 G, 1.26 ERA, 18-4, 3 ND.

That's good luck for Welch, and bad for Clemens. K/BB was 166/43 for Clemens, and 89/41 for Welch. 38 XBH for Clemens and 36 for Welch.

Fair thing here is to say Clemens is 4 games ahead of Welch. Say, make Welch 17-2, 2 ND and Clemens 21-2, 2 ND.

At the 3+ ER:
Welch: 14 G, 5.97 ERA, 8-5, 1 ND
Clemens: 6 G, 5.55 ERA, 3-2, 1 ND

Fair thing here is to make Welch 5-8, 1 ND and Clemens 2-3, 1 ND (i.e., give them both a .400 record)

Total gives me 22-10, 3 ND for Welch and 23-5, 3 ND for Clemens.

These quick win-loss records reflect more accurately the contributions of Welch and Clemens.

The "system" of assigning wins and losses by MLB is not a good one, and certainly not one that would be created, if starting with a clean slate.

6.  As for Stewart, 10 G at 3+ ER (ERA of 7.54), 0-8, 2 ND. This includes a 10 inning, 3 run game.

27 G at 2- ER (ERA of 1.13), 22-3, 2 ND.

Let's make it 23-2, 2 ND, in-keeping with Clemens/Welch. And make the bad games as 1-7,2 ND. Total is 24-9, 4 ND.

Final tally:
Clemens, 23-5
Stewart, 24-9
Welch, 22-10

Stewart had a heckuva lot of great games, but also a disproportionate number of poor games. Welch was pretty lucky.

7.  I agree that both Clemens and Stewart were better pitchers than Welch.

I also agree that assigning wins and losses they way they do is not ideal, and you wouldn't do it that way if you were starting from scratch.

But we're not starting from scratch. If you want, go ahead and take your imaginary, more ideal W-L record and give Clemens an imaginary, more ideal Cy Young award.

I'll take the W-L record from the imperfect world that people actually live in and give Bob Welch a real Cy Young award based on the criteria that real imperfect people actually use, and we can both be happy.

8.  Ken, why not create a W-L record from the imperfect world that we all live in, and give those pitchers the real Cy Young award? What you are saying is that you would rather base it on a W-L criteria that was invented 100 years ago, rather than creating a better W-L criteria today.

I still see no reason to pretend that Welch was responsible for 27 wins. He was responsible for PART of that. Clemens was responsible for part of 21. It's just that Clemens was responsible for a greater part.

I'd hate to talk to you about Nolan Ryan's 8-14.

9.  If you want to talk about reality, of the 21 games that Clemens was assigned a win, I count 12 where he allowed 0 or 1 earned run. So, I give him almost full credit for winning there. 8 where he allowed 2 ER. So, give him half a win there. 3 games where he allowed 3 or 4 ER. So, give him a smidge there. That's 17 wins that he contributed out of those 21.

In his 6 losses, he allowed 1 ER twice (no contibution there), 2 ER twice (give him half a loss there), and 3+ ER twice (give him a loss there). That's 3 losses out of 6 that you attribute to Clemens.

Doing the same with Welch, he gets 20 wins. In his 6 losses, 5 are at his feet.

Incredibly, all of Stewarts wins he allowed 2 ER or less. 19 of the 22 wins he contributed to. He gets 8 of the 11 losses.

So, Clemens contibuted 17-3 of that 21-6. Welch contibuted 20-5 of that 27-6. Stewart 19-8 of the 22-11.

Welch had a good year, but nothing historic. Welch's A's were more historic.