Lurking in the background of any discussion of individual baseball awards is the question of their relevance. Whether they “matter” or not misses the point, though — they’re fun to talk about. Fans, writers, commentators, and even players (witness David Ortiz’s recent outburst) love to ruminate on who deserves what at the end of the season.

When discussing individual award possibilities — especially with two weeks to go in the season — it is important to differentiate exactly what you’re addressing. The conversation should always be broken into two parts:

Who should win? — The player with the best credentials on a wide range of qualitative measures independent of team input or home park advantages/ disadvantages.

Who will win? — The player with the kind of credentials voters usually go for.

Sometimes it is the same man. Other times, depending on a variety of factors, it is not.With Johan Santana of the Twins already clearing space on his mantel for the American League Cy Young Award, we’ll look at the National League contest today.

Who should win?

Before we discuss who should win, we should probably talk about who shouldn’t. There are no relievers who are even close to being good candidates this year. This should come as no surprise. Except in the most extreme cases, relievers do not make good Cy Young Award candidates. Looking at the 1,000 highest VORP (Value Over Replacement Level) scores by pitchers in the expansion era, only 21 were posted by pure relievers. Not surprisingly, no reliever since 1960 has ever had the best VORP in any season.

We should also mention two starting pitchers who shouldn’t win — but not for a lack of trying. They’ve pitched about as well as anybody in the league, just not long enough to be among the frontrunners. Roger Clemens of Houston has pitched extremely well since showing up for work in June. Likewise Anibal Sanchez. Since handcuffing the home team at Yankee Stadium on June 25, he’s had 12 quality starts in 15 tries for the Marlins. Had either been around for the whole year they’d be right in the middle of this discussion.

Looking at the credentials of the starting pitchers who have had a full complement of work, most lists are topped by four men: Chris Carpenter (St. Louis), Bronson Arroyo (Cincinnati),Brandon Webb (Arizona),and Roy Oswalt (Houston). Here are their league rankings in three categories that take into account ballpark factors and leave out the input of their teammates.

In other words, these numbers answer this question: All other things being equal, who pitched best? The first category describes the winning percentage they would have based on how often a pitcher with the same innings pitched and runs allowed in each individual game earned a win or loss historically in the modern era (1972–present). The only other pitcher that ranks among them in this category is Josh Johnson of the Marlins.The second category is their Value Over Replacement Level.

The last category is Average Game Score, a system devised by Bill James to assess the quality of individual starts. For each game, a pitcher starts at 50 points and then gains points for strikeouts and outs recorded and has points subtracted for hits, walks, and runs allowed. For instance, Randy Johnson’s perfect game in 2004 scored 100 while a typical Jose Lima start of the last two years has averaged a 34.

In terms of pure quality starts, independent of offensive support from their teammates, Carpenter has a slight edge on the competition at this point. He’s not far enough ahead that a few great outings by his pursuers or a few slipups on his part won’t unseat him, though.

Who will win?

The things that voters like, such as 20-game winners, are not readily apparent this year. A situation such as that which took place in the American League voting last year is not present in the NL in 2006. With no 20-game winner like the undeserving 2005 winner Bartolo Colon on which to hang their hats, voters are going to have to work around not having a big number. True, if Webb could win his remaining four starts, they’d have a 19-game winner to champion, which, one supposes, is the next-best thing to a 20-game winner. The fact that Webb’s team took the earliest powder in the various pennant races might work to his disadvantage.

As for that question — the relevance of playing for a contender — there is so little separating the contenders and non-contenders in the National League this season (the Mets aside, who don’t have a strong candidate anyway) that it should be moot.

Examining the more traditional statistics, we find that strikeouts are not going to be a factor in determining the outcome:

Again, this is an extremely close contest. Based on the things voters usually like, it would appear to be a battle between Carpenter and Webb with Carpenter presently holding a slight edge. It will ultimately come down to what they do in the remaining fraction of the season. In this regard, the fact that Carpenter is playoff-bound could actually work against him. Cardinal manager Tony La Russa’s desire to set his rotation properly for the postseason might cost Carpenter a start he desperately needs to lock down the Cy Young Award. Webb’s team, with nothing to play for, will be throwing him out there in turn.

Which brings us right back to the relevance question.

Mr. Baker is a writer for Baseball Prospectus. For more state-of-the-art analysis, visit baseballprospectus.com.