I’ve yet to write much about the award races this season, save for some notes about Barry Bonds a couple of months back and a column about the NL Cy Young. It’s a very interesting season in that all six major awards are under dispute, whether warranted or not, and in some cases there’s considerable doubt about who the front-runner should be, even with less than three weeks to go in the season.

Because I got an e-mail this week dissecting the AL Cy Young situation, I’ll take a stab at that one today.

When it comes to picking the best pitchers in the league. I look at two things: how much did you pitch, and how well did you keep runs off the board? I disregard the accounting categories of “wins,” and “losses” because the statistics are misleading, a relic of a time when complete games accounted for nearly 100% of all starts and it actually made sense to assign whole wins and losses to starting pitchers.

With apologies to Mark Mulder, here are the five candidates for the AL Cy Young Award.

                  SNPct.   SNWAR    VORP*   ERA     IP     SO
Esteban Loaiza     .727      6.5    72.1   2.73   204.2   185
Tim Hudson         .691      6.4    66.9   2.61   221.0   144
Pedro Martinez     .761      6.2    61.7   2.36   167.2   187
Roy Halladay       .618      4.9    54.0   3.30   243.0   180
Mike Mussina       .643      4.5    52.9   3.24   197.1   183

(The best reliever in the league is Shigetoshi Hasegawa, who probably has an argument for down-ballot votes but not for top honors.)

Innings and runs allowed are how I judge a starting pitcher, using both the raw data and the Support-Neutral metrics. After that, I’ll consider strikeouts, because they reflect how much of the work the pitcher did on his own. VORP serves largely as a sanity check, which is ironic when you consider that Keith Woolner’s nickname is “5150.”

Based on these stats, Esteban Loaiza would seem to have a edge on the field. He leads in the advanced metrics, and his raw stats match up well with those of Tim Hudson and Pedro Martinez. His innings edge on Martinez is significant, and the ERA difference between him and Hudson hides an unearned-runs gap between the two (two for Loaiza, 11 for Hudson). Roy Halladay has a massive innings lead on the field, but lacks the SN or run-prevention stats. Mike Mussina is comparable to Halladay, less 50 innings.

I mentioned that it was a reader e-mail that sparked this column. Anthony Passaretti raised a good point about the pitchers in this race: they’ve faced considerably different pools of hitters. We know this because Keith Woolner created the Pitchers’ Quality of Batters Faced report. Here’s how the front-runners stack up:

                      AVG   OBP   SLG
Esteban Loaiza       .259  .325  .409
Tim Hudson           .265  .335  .419
Pedro Martinez       .262  .331  .416
Roy Halladay         .267  .339  .435
Mike Mussina         .274  .344  .443

Now, we should have seen this coming. It seemed like every week for a month Loaiza was throwing eight shutout innings against the Tigers. For the season, Loaiza has thrown 44 2/3 innings against the Bengals, about 22% of his total workload. He also caught the Dodgers, Padres (when they were bad) and Cubs in interleague play, all of whom are in the lower half of the NL in Equivalent Average.

Two years ago, Joe Mays wormed his way to the top of the Support-Neutral lists in similar fashion, throwing more than a third of his innings against the Tigers and Royals. Loaiza’s situation isn’t quite so extreme, but as you can see, his superior front-line statistics have been compiled against lesser hitters than everyone else has faced, and considerably lesser ones than what Halladay and Mussina have opposed.

Just for snicks, I dug through the stats to find the full-time players whose performance best matched these lines:

                      AVG   OBP   SLG     Player          EqA  EqR/450
Esteban Loaiza       .259  .325  .409     Todd Walker    .256    74.5
Tim Hudson           .265  .335  .419     Rocco Baldelli .263    79.5
Pedro Martinez       .262  .331  .416     Rocco Baldelli .263    79.5
Roy Halladay         .267  .339  .435     Wil Cordero    .262    79.4
Mike Mussina         .274  .344  .443     Joe Randa      .261    78.4

*EqR/500: EqR normalized to 450 outs

Actually, that’s not as instructive as I thought it would be. The difference between these lines appears to be less than ten runs over a full season. However, park effects and basestealing confuse the issue when it comes to comparing these hitters using EqA, so take those figures with a grain of salt.

Regardless, it is clear that Loaiza has faced a lesser set of hitters in 2003 than the other four pitchers have. Is that enough to push Hudson, Martinez or Halladay ahead of him? It seems like the spread here is less than a full win, but then again, so is the gap between Loaiza and his two closest pursuers in SNWAR and VORP. It is clear that this is a factor we have to consider in all the award voting; the unbalanced schedule requires that we use greater vigilance in player evaluation.

For what it’s worth, Hudson has worked the whole year in front of the best defense in the league, and Loaiza for the fourth-best, according to Defensive Efficiency. The Red Sox, Blue Jays and Yankees are 11th through 13th in the AL. Hudson, in particular, has gotten great defensive support: he’s allowing just a .240 batting average on balls in play. Mussina’s defensive support has been better than that given to other Yankee pitchers as well.

I’m just glad there’s two weeks to go, because I don’t know which of these pitchers is most deserving of the honor. Loaiza has the metrics, but Halladay has the innings and picks up ground when you start looking at opposition and defense. Everyone else kind of falls in between.

This isn’t going to be decided until the last starts of the season for all involved, but right now, I’d probably list the candidates as Loaiza, Halladay, Mussina, Martinez, and Hudson, with the first two separated from the next two, and those two separated from Hudson.