The Clemens/Carpenter Debate

By | August 16, 2005

I just finished reading the latest chat with Insider Buster Olney (here’s the link, for those of you who have Insider accounts). Throughout the chat, he printed a lot of comments about the Roger Clemens vs. Chris Carpenter debate. I noticed that most of the pro-Carpenter arguments seemed to really be anti-Clemens arguments: “If the stats were reversed, you would all say Clemens deserved it.” “Clemens didn’t deserve it based on stats last year, but he won it anyway.” And so on an so forth.

As usual, I have some opinions on this topic, and like most debates like this, I think a lot of people are using emotion instead of data to make their case. I will now address several of the arguments I have read today:

As of today, August 16, Chris Carpenter is 17-4 with three no-decisions, and Roger Clemens is 11-4 with seven no-decisions. Each has started 24 games, which makes it pretty easy to compare them.

As we all know, baseball is a team sport, and while the pitcher gets credit for wins and losses, it is often someone else, whether a hitter or a fielder or a manager, who deserves some or all of the credit/blame. That’s why the Earned Run Average statistic exists: to show what the pitcher himself does. And that is why ERA is still the most accurate reflection of how well a pitcher has pitched.

Right now, Roger Clemens has a 1.32 ERA for the season, which would be the best ERA since Bob Gibson’s 1.12 in 1968. Chris Carpenter has a 2.25 ERA, which is also magnificent. But no matter how you slice it, the ERA difference tells us one thing: Clemens gives up nearly one entire run fewer than Carpenter every nine innings.

Here’s one of the arguments that I have heard a few times, taken from this Buster Olney chat:

Joe (Billings, MT): If you take away one bad outing in April, Carpenter’s ERA is 1.88 and he has won 6 more games than Clemens. Seems obvious that Carpenter is the Cy Young winner and Clemens is an excellent 2nd place choice.

I am not sure how this argument became so popular, because it seems to me that any argument that starts with “If…” needs to be followed up with my parents’ favorite saying: “If my daddy had wheels, he might be a wagon.” It is absolutely true that IF Chris Carpenter hadn’t had a terrible game in April, his ERA would be within half a run of Clemens’ ERA. But if you are willing to take away Carpenter’s one bad outing, don’t you also, out of fairness, have to take away Clemens’ worst outing? Well, when you do that, all of a sudden his ERA has gone from 1.32 to 1.13!

Joe from Billings also mentions the evidence most often cited in this argument: Carpenter has won 17 games, and Clemens has only won 11. As I mentioned before, you can’t give all the credit or blame to a pitcher for the outcome of the games he pitches. Here’s a way I came up with to compare these two on somewhat equal ground: In the games Chris Carpenter has started, the Cardinals have scored an average of 5.17 runs (124 runs in 24 starts), while the Astros have averaged 3.67 runs in Clemens’ games (88 in 24). What I have done is figured out what each man’s record would be, based on how well he pitched that game, if his team had scored that average. So basically, for Carpenter, any game where his Game ERA (his ERA just for that game) is under 5.17 is a win, and anything over 5.17 is a loss; for Clemens, the same thing, but his target ERA is 3.67.

Using this formula, Clemens and Carpenter both come out with a record of 22-2. There is absolutely no question that if both men were 22-2 right now, with Clemens’ ERA nearly a run lower, this debate would not be happening.

But wait, it gets better. What if the roles were reversed? What if Clemens got Carpenter’s run support, and Carpenter got Clemens’ run support? Well, in that case, Carpenter would be 17-7, which is still a pretty darn good season. Clemens? He would be 24-0.

In 17 of Clemens’ 24 starts, he has given up either one earned run or zero. Only twice has he given up more than two. Carpenter has given up one or zero earned runs in 12 of his 24 starts, with more than two eight times.

Here’s a big argument that combines a lot of the pro-Carpenter points:

Chris (St Louis): Other than the ERA I don’t think the CY Young is close. Carpenter has 6 complete games, 4 shutouts, more strikeouts, 7 more wins and is on a team that has lead wire to wire for the division. Also, its not like Carpenter is getting 7 runs a game when he pitches. Oh yeah, he also beat Clemens head to head…I think the race is actually over. By the way, why does the criteria change year to year? Last year Randy Johnson had better stats than Clemens except wins and Clemens won the award, this year the way we are judging is totally different.

Let’s address these point by point, I guess:

  • “Carpenter has 6 complete games”
    Yes, Chris Carpenter has six complete games, and Roger Clemens has zero. Well, let’s remember that this is the National League, where a pitcher often comes out of a game for reasons completely unrelated to his pitching. I think this might be the most underlooked benefit to a pitcher on a team with a good offense: if a pitcher comes up in the 8th inning with a four-run lead and he is pitching a good game, he will stay in the game. If a pitcher comes up in the 8th inning of a scoreless tie, he will be pulled for a pinch hitter every time, regardless of the fact that he is pitching a shutout.

    So how many times has Roger Clemens been pulled for a pinch hitter in the 7th inning or later in a game that was within two runs? Exactly ten. Ten times! Only once out of those ten was his pitch count at a point where he would probably have been hit for anyway (he had thrown 126 pitches when he was hit for in the 8th inning of a 0-0 game on April 23). So there were nine times when Clemens’ bid for a complete game was thwarted by his team’s inability to give him a comfortable lead.

    Of course, to be fair, we need to look at how many times Carpenter has been pulled in that same situation. There have been two times when Carpenter was pulled for a pinch hitter in the 7th or later in a game that was within two runs. He was at 90 and 112 pitches in those two games, so it is reasonable that in a less-close game, he would have been allowed to go for the complete game.

    Carpenter’s six complete games have had an average margin of victory of 4.5 runs, with only one decided by two runs. If this were the American League, it is reasonable to guess that he would have eight complete games. Of course, if it were the American League, Clemens could have nine complete games. So, it appears that complete games aren’t as much of a pitcher stat as they are a team stat.

  • “Carpenter has 4 shutouts”
    Clemens has been pulled for a pinch hitter three times when he was in the process of pitching a shutout. Carpenter: zero. So again, with a little more run support, this statistical category could be a lot closer than it is. Shutouts are the fraternal twin of complete games, and are just as dependent on team support.
  • “Carpenter has more strikeouts”
    Yes, Carpenter has 169 strikeouts to Clemens’ 148. When you look at the K/9 numbers, though, it is quite a bit closer: Carpenter has 8.3 to Clemens’ 8.1. And as we have already established that Clemens has had several innings stolen by an anemic offense, the ratio approach is the only fair way to do it.

    So yes, Carpenter has 0.2 more strikeouts every nine innings, but that stat certainly isn’t enough to give him the Cy Young Award. Especially when you consider that Nolan Ryan never won a Cy Young.

  • “Carpenter is on a team that has lead wire to wire for the division”
    Yes, Carpenter is on a great team. I hope they win the World Series, because that is the award for the best team. Given the fact that Carpenter’s ERA was still above 4.00 after his ninth start of the season, I question how responsible he is for the “wire to wire” aspect of that statement. I think it is apparent that the Cardinals are at least as responsible for Carpenter’s success as Carpenter is for the Cardinals’ success.
  • “Also, its not like Carpenter is getting 7 runs a game when he pitches”
    First of all: LINK
    Okay, now, Chris is right. Carpenter ISN’T getting seven runs a game when he pitches. But he IS getting nearly a run-and-a-half more than Clemens is. And he DID get 6.44 runs per game during a mediocre first nine starts that somehow saw him end up at 6-2 with a 4.07 ERA.
  • “Oh yeah, he also beat Clemens head to head”
    Did they play a game of one-on-one that I didn’t hear about? Yes, the Cardinals beat the Astors on July 17. Carpenter pitched a complete game shutout; Clemens gave up one earned run in seven innings. I must be missing some sort of significance here, because I see this and I think, “Yeah, Carpenter was slightly better that game.” I don’t look at that and think, “Well, there is proof that Carpenter is better.”

    Let’s look at it this way: in 1988, Orel Hershiser won the Cy Young Award with a 23-8 record. He was beat, head to head, by:

    • John Smiley
    • John Dopson
    • Ron Robinson
    • Bob Walk
    • Doug Drabek
    • Jim Deshaies
    • Kelly Downs
    • Bob Ojeda

    What does that mean? Absolutely nothing, just like Carpenter beating Clemens head to head means absolutely nothing.

  • “Last year Randy Johnson had better stats than Clemens except wins and Clemens won the award, this year the way we are judging is totally different”
    It seems that the same people who are saying people favor Clemens because of his total career and aren’t basing it on just this season are the same ones who want to make the argument that Clemens didn’t deserve it last year so he shouldn’t get it this year. It seems kind of backwards to me.

Chris Carpenter is having a great season, and I am happy for him. He seems like a classy guy, humble, a good teammate, etc., and in most seasons he would be the Cy Young favorite. But when you look past the win/loss record, for which a pitcher is not completely responsible, you see that Roger Clemens is having one of the best seasons we have seen in the past 50 years. Yes, it adds a little bit that he has had such a great career. Yes, the fact that he could win two Cy Youngs after retiring is awe-inspiring. But if someone put the season stats of these two men in front of me with no names, I would pick the man with Clemens’ stats for Cy Young every time.

Leave a Reply

Your email address will not be published. Required fields are marked *