K/BB ratio -- for many statheads, that one number tells you most of what you need to know about a pitcher.  The higher the K/BB, the lower the ERA, and vice versa.  But every year, there are a number of starters who post good ERAs with crappy K/BB ratios, and fans of those teams come up with moderately rational reasons why they can continue to beat the K/BB rule in the future.  Here are three of those guys this year:

 xFIP is expected ERA based on K/BB ratio and GB% (and nothing else).  These three pitchers seem to have greatly overachieved.  But how sure of that are we?  More thorough studies have been done, but here's another one that hopefully makes the point that a pitcher succeeding with a low K/BB is much more likely to be getting lucky than to have beaten the system.

I grabbed all pitcher-seasons from 2004 through 2008 (Hardball Times only goes back through 2004) with at least 120 innings pitched and divided them up into three groups based on K/BB (lower than 1.5, between 1.5 and 2.5, and above 2.5.)  Here is each group's average ERA and percentage of pitcher-seasons with an ERA under 4.50:

It's obvious that a low K/BB is not a good recipe for success.  But the question remains, can certain pitchers find ways to be exceptions to the rule?  Let's look at all the pitchers who have posted an ERA under 4.50 with a low K/BB from 2004 through 2007, who also pitched at least 80 innings the following year.  That should show us the pitchers who proved year one wasn't a fluke:

Wait?  That's it?  Yep -- only four starters have backed up a sub-4.50 ERA season with another one, given a low K/BB the first year.  As a group, the pitchers in the table saw their ERAs rise from 3.82 to 4.71 from the first year to the second.  And that change could easily have been predicted -- their average expected ERA (xFIP) the first year was about 4.76.

But ok, maybe those four pitchers who followed up their "flukey" year with another one actually can beat the system.  (Remember, this is four players out of 89.)  Let's take a look at them individually:

Brandon Webb 2004

Webb's a crazy groundball pitcher (63%!), which means he doesn't give up a lot of homeruns -- his ERA was only half a run lower than expected in 2004.  And in 2005, his K/BB ratio skyrocketed all the way to 2.9 and his ERA was actually a half run worse than expected.

Chien Ming Wang 2006

Wang is also a crazy groundball pitcher and outperformed his expected ERA by .80 runs in 2006.  In 2007, his K/BB rose a bit to 1.8.

Aaron Cook 2007

Cook (stop me if this sounds familiar) is an extreme groundball pitcher.  His 2008 K/BB is up to 2.0 and he's only outperforming his expected ERA by .30 runs.

Victor Zambrano 2004

Zambrano bumped his K/BB up from 1.1 to 1.5 in 2005, but to be fair, he did outperform his K/BB again.

In summary, it's extremely difficult to post an ERA under 4.50 with a K/BB under 1.5.  Of those pitchers who managed to do it between 2004 and 2007, only four repeated the following year.  Of those four, three are extreme groundball pitchers -- who also increased their K/BB to varying degrees -- and Victor Zambrano.  And the only good aspect of having the next Victor Zambrano on your team is that maybe he'll be traded for the next Scott Kazmir.

Finally, let's go back and take a look at the three starters who, in 2008, are sporting ERAs under 4.50 with a K/BB under 1.5.  What are the chances that Edwin Jackson, Dana Eveland, and Greg Smith repeat their low ERAs 2009 with similar K/BB ratios?  All three pitch in front of very good defenses (Tampa Bay and Oakland), and the two Oakland pitchers play home games in a friendly park for pitchers. But otherwise, Eveland is the only one with a high GB-rate (and he's the one currently outperforming his xFIP the least.)  Jackson and Smith should see quite the regression next year.  Of course, all three guys have room to improve their skills and therefore their K/BB ratios, but being able to beat the system and becoming a better pitcher are two separate things.