There were three legitimate candidates for the National League Cy Young Award this year in Clayton Kershaw, Roy Halladay and Cliff Lee, and I spent just about all of my time on this ballot weighing the performances of those three pitchers; there were other pitchers who had great years, but I don't see a good argument for any top three beyond the pitchers I just named. A quick glance at the most superficial of stats shows just how tough filling out this ballot was going to be:
National League Cy Young Award candidates
IP R RA
Pitcher A 233.1 66 2.545
Pitcher B 233.2 65 2.503
Pitcher C 232.2 66 2.553
Even if we use ERA instead of RA, the difference between the lowest and the highest in this troika is just 0.12, equivalent to three earned runs allowed over the course of the entire season. There's just nothing here to support the mainstream narrative that Kershaw ran away with this award (although given that narrative, I'm not surprised he won). Kershaw struck out 28 more batters than Halladay did in the same number of innings, which is great. He also walked 21 more (exclusive of intentional walks) and gave up five more home runs, which is why their run averages were so similar, and why advanced metrics rank these players in a surprisingly different order.
The biggest separator between the three names for me, before we even get to advanced metrics, is strength of schedule. Kershaw made six starts this year against the San Francisco Giants, throwing 42 innings and allowing just 6 runs, and was almost as effective against the San Diego Padres, with 25 innings and 5 runs allowed. Those are two poor offensive teams -- they were last and second-to-last in runs scored and, not coincidentally, OBP -- playing in extreme pitchers' parks. Halladay never faced the Giants and threw 16 innings against San Diego; Lee made one start against each club for 13 innings, shutting out the Giants but giving up 10 hits and 5 runs in four innings to the Padres. Kershaw can't control his schedule, of course, but it's important to understand the magnitude of the challenge he faced (or didn't face) in posting his final stat line. If Halladay and Lee pitched just as well, but faced tougher competition, shouldn't they be rewarded for it on my ballot?
Once we move to advanced stats, the separation becomes even clearer. FanGraphs' version of wins above replacement, which normalizes batting averages on balls in play, has Halladay well ahead of anyone else in the NL at 8.2 WAR, nearly a win and a half ahead of Kershaw and Lee, who are separated by just 0.1 (effectively tied). The idea behind such stats, and the reason I like to use them when evaluating pitchers for the purposes of voting on the Cy Young Award, is that they strip out the aspects of pitching over which the pitcher has either no or limited control, events like his defense making or not making plays behind him and the bullpen stranding runners he leaves on base or allowing them to score. Halladay also led the NL in Baseball-Reference's version of WAR, which doesn't adjust for batting average on balls in play; his lead over Kershaw there was 0.4, with Lee once again 0.1 behind the Dodgers lefty.
We are so accustomed to giving all of the credit for run prevention (or blame for non-prevention) to the pitcher, rather than attempting to divide that credit between the pitcher and the defense. Our current tools are not perfect, but they are better than pitcher wins or saves or ERA for giving us a clearer picture of the value a pitcher contributed to his team.
With all of those considerations in mind, I had Halladay (pitcher B) as the easy choice for the top spot on my ballot. Lee (pitcher C) versus Kershaw (pitcher A) was a toss-up, and I went with the pitcher who faced tougher opposition, Lee, in the second spot, with Kershaw third. Matt Cain and Cole Hamels rounded out the ballot in spots four and five.