Why I Think the NL Cy Voters Got It Wrong

Don't Misunderstand Me: I'd Start My Expansion Team With Any of These Guys

As late as September 12, I had been on record as saying that I thought Chris Carpenter should win the Cy Young.

But by September 27, I was sure that Carpenter's spectacular September nosedive had sealed the deal for Dontrelle Willis. Now, I see that the writers have picked Carpenter after all, and while it's not a horrible pick (not like Colon over Santana, sheesh!), there's this little yellow man in my head saying that the writers got it wrong.

It's become fashionable these days to say that the games down the stretch don't mean any more than the games in April, and in a certain way, that's absolutely true. But on another level, it totally misses the point. Lose a game when you're three games out on April 27, and you've got 140 games left to try to steal that one back. Lose it when you're three out on September 27, and you've put a serious damper on the team's collective travel plans for mid- and late October.

Let's take a look at our NL Cy Young cast of characters, circa August 15. Rank 'em by wins, or by the length of the names I'll use, what the hell:

On August 15

Pitcher	W - L	ERA	WHIP	IP/GS
Carp	17 - 4	2.25	0.98	7.51
Willis	15 - 8	2.79	1.13	6.86
Clemens	11 - 4	1.32	0.93	6.83

If you all believe in the idea that you should try to finish what you start, you might see how Chris Carpenter made the strongest case back on the Ides of August. And even if you don't, thinking that's why you signed that setup man to the incentive-laden contract, you should be able to recognize that these are three very good pitchers, packed very tightly together.

To my mind, by August 15, Carpenter, Willis, and Clemens had at the very least served notice that one of their trio was the best pitcher in the National League. This despite some interesting performances strung together at various times by Andy Pettitte, John Patterson, and Pedro Martinez.

Certainly, if somebody had been unable to make up their mind about ranking these three on the Cy scale back on August 15, the idea that maybe how the trio did to finish the year could settle the question. After all, they each pitched for teams that had some realistic postseason chances. The idea that the Cy Young winner has a responsibility to win down the stretch for his team, to push 'em over the top, I guess, seems antiquated and old-fashioned these days, when sabermetricians seek to separate individual and team achievement in ways that pretend one has no bearing on the other.

But I guess I'm just old-fashioned. Let's take a look at the contenders down the stretch. This time I think I'll rank 'em in terms of Cy worthiness, 'cause if it was more or less a tie on August 15, it was the games thereafter that broke it, at least to my mind:

 

September 1 - October 2

Pitcher	W - L	ERA	WHIP	IP/GS
Willis	4 - 2	2.68	1.17	7.19
Clemens	2 - 2	4.33	1.63	5.40
Carpenter	2 - 1	5.73	1.49	6.28

Willis was obviously the best of the three during the last month. You might say that Carpenter should have been ranked behind Willis, 'cause of the WHIP and the one fewer loss, but 1) it's not like a 1.49 WHIP is any good, either, and 2) the two victories Clemens did get--against Florida September 14, and against Chicago October 1--were absolutely critical. Each time, you had the feeling if the Astros lost, that was it. And each time the Rocket won. And while Carpenter beat the Mets in early September in what you could argue might have been a game with critical playoff implications, what sticks out is September 30, when Carp was lousy against the Astros. 3) The excuses Carpenter made when interviewed during the playoffs didn't sit well with me, either.

But anyway, forget second place, there you have it: Dontrelle Willis should have won him some Cy Young.

Thank you and Goodnight.

* * * * While I'm pretty sure I've managed to deal with the reasons why Carpenter shouldn't have won, as I was getting ready to post this, it occurred that some still might not buy my Willis over Clemens argument. I just kept hearing these whiny sabermetric voices: "Sure, Willis was best down the stretch, but Clemens' numbers aside from wins were so good for 2/3 of the season that in essence, he couldn't've been caught even after he faltered."

I guess I wasn't gonna get around it. Clemens had the 1.87 ERA and the .198 BAA, those who felt he should've won might say, well, he had so much bad luck. Clemens' run support was terrible.

His run support was terrible, no question. 3.58 runs per 9 innings of Rocketball. Pretty anemic. But if we're gonna give Roger credit for what his teammates didn't do for him, then why not vote for Javy Vazquez? Vazquez, if you can believe it, got even less supoort from his D-Back teammates that Roger got from the 'Stros, PLUS his strikeout to walk ratio was higher, as was his K/9IP.

I guess they call that reductio ad absurdium. We feel bad for Roger (and Javy), but if it's wrong to give the award to Carpenter because he pitches for a team that features Albert Pujols and Jim Edmonds, then it's just as wrong to give it to Clemens because the Astros did their best Hitless Wonders impersonation when he was on the mound.

Let's take a look at won/loss vs. run support:

Clemens
4 - 0 when giving up 0 runs (10 times)    6 ND
6 - 0 when giving up a run   (8 times)    1 ND
2 - 2 when giving up 2 runs  (7 times)    3 ND
0 - 3 when giving up 3 runs  (3 times)    0 ND
1 - 0 when giving up 4 runs  (1 time)     0 ND
0 - 2 when giving up 5 runs  (2 times)    0 ND
0 - 1 when giving up 6 runs  (1 time)     0 ND


Carpenter
7 - 0 when giving up no runs(7 times) 0 ND
6 - 0 when giving up a run  (7 times) 1 ND
2 - 0 when giving up 2 runs (3 times) 1 ND
6 - 2 when giving up 3 runs (8 times) 0 ND
0 - 0 when giving up 4 runs (3 times) 3 ND
0 - 1 when giving up 5 runs (2 times) 1 ND
0 - 0 when giving up 6 runs (1 time)  1 ND
0 - 1 when giving up 8 runs (1 time)  0 ND
0 - 1 when giving up 9 runs (1 time)  0 ND

Willis
8 - 0 when giving up no runs (9 times)  1 ND
6 - 1 when giving up a run   (7 times)  0 ND
5 - 1 when giving up 2 runs  (6 times)  0 ND
2 - 2 when giving up 3 runs  (5 times)  1 ND
1 - 1 when giving up 4 runs  (2 times)  0 ND
0 - 1 when giving up 5 runs  (1 time)   0 ND
0 - 1 when giving up 7 runs  (1 time)   0 ND
0 - 2 when giving up 8 runs  (2 times)  0 ND
0 - 1 when giving up 9 runs  (1 time)   0 ND

Everyone has focussed on the fact that the Astros were shut out in seven of Roger's starts. Some others have expressed the truism as I do: that of the ten times Roger gave up no runs in a start, he grabbed the win in only four of them.

But did you know that Willis gave up no runs nine times? Or that Carpenter did it seven times? Or that Carpenter had more quality starts than Clemens, for that matter? And did you realize that not once did Clemens give up a single run then lose? All the talk of poor support would lead you to believe that Roger was handed losses he didn't deserve, but mostly, he was denied wins.

Clemens had the same record as Carpenter when giving up a single run, and a better one than Willis.

With all the tough losses (five of them, it is true), Clemens was also the only one of the three to pick up a Cheap Win. (Clemens was additionally only third on his own team in Tough Losses). So let's do this: take a look at the one no-decision Dontrelle took when giving up nothing, subtract it from Roger's six. Take the five and add it to 13, then subtract the Cheap Win. You then get an RSAWT^* of 17. You adjust for the biggest inequity, and Clemens still comes four short of Carpenter and five short of Dontrelle, in the most important stat of all.

The fact is, with a huge September, Clemens could have grabbed the Cy Young for his very own. I'll assume that it was because of a hamstring that he didn't, but instead he floundered; up one game, down the next, skipping the third. Willis kept his eye on the prize (for the most part) and did what I feel was necessary to earn the hardware.

----
^*Rastronomicals' Simply Adjusted Win Total