In 2002, Bill James released a simple formula that, up to 2002, had predicted 46-of-51 Cy Young winners correctly. 

This excludes 1969 to 1984, when sportswriters were confused about what to do with the new relief aces, but even including those years James was still 81% accurate overall.

His system also calls surprises correctly, and frequently nails the top 3 dead-on, even when not very predictable.  For example, in 1990 the top 3 vote-getters were Drabek, Ramon Martinez, and Frank Viola; James' formula called those three exactly.

..............

His system is:

W x 6

Minus L x 2

+ K / 12

+ SV x 2.5

+ ShO

+ Runs Saved (reflecting ERA and volume IP in correct proportion)

+12 points pitching for first-place team.

.

=== Zen Buddhism Dept. ===

Zen, of course, being the art of the simple, clear, and important insight into the nature of the universe.  :- )

You're not going to be able to create anything simpler than James' system.  Its parts, like a mousetrap's, are irreducible.  You're not going to remove strikeouts, use one less moving part, and be as accurate.

Saberdudes sometimes try to introduce 40-variable fomulas (including, for instance, xERA or FIP which contain many variables) to try to beat James by a percentage point or two. 

James, however, has a fine feel for the most important factors in a discussion, which are the above.   His formula, in essence, psychoanalyzes sportswriters and diagnoses what has been important to them.  Besides access to Erik Bedard, of course.

The purpose of the formula, as James states, is not to call the Winner after the season.  It is to track the race during the season, to "make a reasonable conclusion about who is ahead in the race, how far ahead they are, and what the other pitchers have to do to catch up."

............

I was wondering whether to spend a half-hour calculating the top guys in 2009 per his formula.  I didn't.  Guess why?  MLB.com has it running daily!

.

=== Pole Position Dept. ===

In real life, the field is tightly packed as it roars around the final turn.  In real life, there isn't a lot of difference this year between Verlander, and Felix, and Greinke, and Sabathia, and Halladay.

James' formula reflects this beautifully:  the separation between the top six guys is small but distinct.  Mariano Rivera is a viable candidate for the Cy, because he's "leading his team to the pennant."

The top guys are, with their James Cy points:

181 - Felix

176 - Greinke

169 - Sabathia

163 - Rivera

162 - Halladay

160 - Verlander

Notice that, since wins are worth 6 and losses are -2, we are in a situation where a final weekend could decide the voting in James' formula.  This is another charming parallel to real life.

.

=== Why Felix?, Dept. ===

Cy Young voters like wins, they like 1st-place players, they know about strikeouts.

But what we all have been missing is that sportswriters absolutely love low loss totals in their Cy Young winners.

Felix has a mere 5 losses all season -- despite terrible run support.   Everybody else has at least eight.   Greinke and CC have 8 ... Verlander has 9 ... Halladay has 10.  Felix has only five.

The 18-5 record conveys the aura of perfection, of super-humanness, of transcending the people around him. 

We know that W-L isn't perfect, but the fact is that history's gaudy W-L records won a lot of Cy's.   We all remember Zito's 23-5, Pedro's 23-4, Maddux' 19-2, and Welch's 27-6.

18-5 for Felix?  What was Tim Lincecum's record last year?  :- )

.

=== D-O-V Merit Award ===

It's popular to say that W, L, R, RBI, etc mean absolutely nothing.   They can be misleading, but it's overstating the case to say that they mean nothing.

Supposing that you snuck a peek at Biff's Sports Almanac 2000-2050 and found a line saying 2010 M. CARP, Sea - 147 RBI. 

That would mean "absolutely nothing" to you, except that he would play all year?  

Supposing Biff's almanac revealed that Luke French was going to go 22-4 next year.  That would mean "absolutely nothing" about whether French was going to pitch well?   

.

=== One (1) Off Start in Four+ Months ===

FIP is cool, but baseball is too complicated for one stat.  A man could allow 5 runs, 0 runs, 5 runs, 0 runs, all season long, and finish with the same FIP as a man allowing 2.5 all the time.  The man allowing 2.5 would win far more games.

..................

Felix Hernandez has had one (1) off start since May 19.   He has had precisely one game in which he gave up > 3 runs since then -- and that's the primary reason he's never lost.

The King has been awfully tough to beat.  James' system points us to the key factor in this race -- his start-to-start consistency.

Felix has been invincible.  I think that's what you want in your ace.

Cheers,

Dr. D