The 2005 AL Cy Young Award

True to the hint dropped at the end of my last post, I'm going to explore the 2005 AL Cy Young Award, but first let me be clear about my purpose.

While I believe that Johan Santana and Mariano Rivera were clearly the two best pitchers in the league that year, I'm not going to argue that one of them should have won the award. I recognize that each of them faced a significant challenge to winning the award, challenges that the BBWAA rarely works hard enough to overlook. In Santana's case, his team's general mediocrity suppressed his win total, and in Rivera's case, his role as a reliever kept him from throwing very many innings. I could go chapter and verse on why either man was a better pitcher in 2005 than anyone else in the league, but that's been done before by many, and I'm not in the mood to do it again.

Instead, I want to have a discussion of how the BBWAA evaluates candidates for the Cy Young in general, and I think I've identified the perfect candidates for framing this discussion.

Bartolo Colon was the winner of the American League Cy Young Award in 2005, receiving 17 of the 28 possible first-place votes. This award was based upon Colon's 21-8 record and 3.48 ERA for the AL West champion Los Angeles Angels of Anaheim. His teammate, John Lackey, posted a record of 14-5 and a similar ERA of 3.44. He didn't receive a single Cy Young vote, but when his performance for the season is compared to Colon's, I think the similarities are noteworthy.

Both players started 33 games, and the Angels went 22-11 in both Colon's starts and Lackey's starts, so clearly each man gave his team an equal chance to win. Colon was a bit more durable and a bit more efficient with his pitches than Lackey was. He walked 28 fewer hitters and gave up fewer hits per inning, while hitting fewer batters and throwing considerably fewer wild pitches. All of these results generally tended to keep his pitch counts down and extend the time he could spend in each game. Consequently, Colon threw almost 14 more innings than Lackey did. Colon also threw three more quality starts than did Lackey, 21-18. All of this led to Colon getting the decision in 29 of his 33 starts, nearly 88%, compared to just 19 for Lackey, or 58%. Much of that is due to Colon's efficiency and durability, and he should absolutely get credit for that. Taking a look at the WARP3 marks (7.5 for Colon, 6.5 for Lackey) and Win Shares (19 for Colon, 17 for Lackey) for each shows us that he does. I will not try to argue that Colon wasn't slightly better than John Lackey in 2005, because clearly he was.

That said, there are quite a few factors in Lackey's favor that narrow the gap between them significantly:

• Lackey faced tougher competition. Eighteen of his 33 starts came against teams with records of .500 or better. This compares to 15 for Colon, who racked up 11 of his 21 wins against sub-.500 teams, while Lackey posted 11 of his 14 wins against winning ballclubs.
• Lackey pitched more of his games on the road. While Colon got to start 18 of his games in Anaheim, Lackey had to start 19 games on the road.
• Lackey enjoyed considerably less run support than Colon did. On average, the Angels scored two-thirds of a run more in each of Colon's starts than in Lackey's. Seven different times the Angels scored double-digit runs for Colon, compared to just once for Lackey. This contributed to the fact that Lackey turned in six quality starts in which he got no decision, compared to just three for Colon.
• On average, the Angels allowed fewer total runs in Lackey's starts, committing fewer errors and allowing fewer unearned runs, a traditional hallmark of a more engaged defense. In other words, his teammates seemed more interested in playing behind him.
• Lackey allowed half as many home runs, just 13 to Colon's 26, while striking out batters at a much higher rate, 8.6 per nine innings to just 6.3 for Colon, indicating that on an inning-for-inning basis, Lackey demonstrated more dominance over his opponents than Colon did.
• The Angels' bullpen blew three saves after Lackey left the game, cutting his possible win total by three, while they didn't blow a single lead for Colon all season.

When all of these factors are combined, and the respective records of the two pitchers are neutralized (according to baseball-reference.com) to account for home-road discrepancies, run support, etc., we find that their marks aren't terribly different:

Colon: 14-10, 3.33 ERA, 224.3 innings, 158 strikeouts, 42 walks
Lackey: 14-9, 3.21 ERA, 210.3 innings, 200 strikeouts, 70 walks

In other words, Lackey's neutralized record is very close to the record he actual posted, while Colon's makes it clear that his final numbers were largely the result of fortunate circumstances beyond his control.

Now, even presuming that Colon should get an extra couple of wins due to durability issues, the question becomes this; Why did Bartolo Colon win the Cy Young Award in 2005 when a teammate with an extremely similar performance didn't receive a single vote? Was Colon so much better than Lackey that he deserved to be labeled the best pitcher in the league while Lackey received no consideration for that title at all? I don't think so.

I think the BBWAA saw that league-leading win total and stopped asking questions. Colon won more games than anyone and he did it for a division winner, so the question of whether or not his performance was really indicative of the best pitching in the league became an open and shut case for most voters. At the same time, Lackey's comparatively pedestrian win total, particularly in the absence of a league leading ERA or strikeout total, eliminated him from consideration for most, if not all, of the writers before they even filled out their ballots.

This is pretty typical for the BBWAA. Time and again, we've seen them focus on one or two key statistics - wins for starters and saves for relievers in the Cy Young voting, RBI for hitters in the MVP voting - and essentially halt their performance analysis at that point. I don't agree with it, but I've come to expect it.

Still, when you've got a situation like Colon and Lackey in 2005, you'd hope the voters would take a bit more time to examine just a couple of basics. If any one of them had bothered to look at them side-by-side, they would have seen two teammates with nearly identical ERAs who started the same number of games for the same team, and the team posted identical 22-11 records in their respective starts. On the field, where all of this is supposed to matter, John Lackey and Bartolo Colon provided essentially equal value to the Angels, with Colon providing a touch more due to his added durability. Wouldn't the voters, if they were really doing their jobs, takes pains to ensure that the voting results reflected this reality? If you really think Colon was the best pitcher in the league, shouldn't Lackey be somewhere down your ballot, probably just one slot below Colon?

Instead, we get the writers sending the tacit message that they really only consider one or two numbers when they cast their votes, whether those are the proper numbers to consider or not. Silly results ensue, leaving anyone who cares about this stuff either scratching their head or, even worse, expressing no surprise at all considering how frequently this kind of thing happens.

For me, I can't help but think that it's a pretty sad commentary on the voting process when it's no longer surprising that the results of the writers' ballots don't match the reality we witness on the field.