Comparing Ranges of K% and BB%

Recently, I’ve begun comparing pitchers based on their K% and BB% going back to 2002, which is when batted ball data begins to become available on FanGraphs. My thought is this: pitchers have pretty good control over their K% and BB%, so those reflect true skill. What I want to see is a range of effectiveness for pitchers in the MLB within a set range of K% and BB% to compare players. I recently looked into John Lamb, who is a pitching prospect in the Royals minor league system. What follows is not a rigorous analysis but just something interesting I observed. 

Lamb used to be a top prospect when the Royals had The Greatest Farm System in the History of Whenever (TGFSITHOW). Injuries and a decline in effectiveness have pushed him far down the list of top prospects, but I was interested anyway. I looked at his minor league K% and BB% numbers to see how he compares to major leaguers with similar rates. In 2013, Lamb had a 19.1% K% and 4.8% BB% in High-A ball in 92.2 innings; this compares favorably. It is very easy to be successful in the MLB with those kinds of rates. However, I don’t believe those rates to be representative, as he is quite old for High-A. 

What is more interesting is looking at his 2011 AA numbers. Lamb threw only 35 innings in AA that year, so his numbers come with the small sample size caveat. His K% was 14.8%, and his BB% was 8.7%. Neither number is very good, but I saw a neat comparison to MLB pitchers in 2013 who had a K% between 14-15% and a BB% between 8-9%. There were 7 such pitchers (min 50 innings), both starters and relievers, whose WAR ranged from 1.1 to -0.6. The GB% is the interesting bit. The highest GB% corresponded with the highest WAR (Samuel Deduno of Minnesota), the 2nd highest GB% corresponded with the 2nd highest WAR (Luis Avilan of Atlanta), and so forth. Only the 5th and 6th ranked GB% didn’t correspond with the 5th and 6th ranked WAR, but in 2013 there was a clear relationship between GB% and WAR among pitchers in the K% and BB% range. In 2013, a higher GB% meant a higher WAR. 

When looking at this relationship with a larger sample of data (not single season data, but career data accrued in 2002-2013), the relationship doesn’t seem to hold as well. The more successful pitchers with this K% and BB% were indeed groundball pitchers, but there were some groundball pitchers who weren’t successful as well. Successful in this K% and BB% range is a relative term; the pitcher with the most WAR in this sample, Roberto Hernandez, is mediocre at best, but he has had a fairly long career with over 1,000 innings. In general, there are not many pitchers with this K% and BB% profile, and the pitchers who are there are not very good. If John Lamb is to see the majors, he has to be an extreme groundball pitcher (which he doesn't appear to be), or he has to have better control.