|Evaluating Pitchers: DIPS, Walks, and Strikeouts|
Written by Brian James Oak (Contact & Archive) on August 07, 2007
The first thing to know about the current state of pitching projections is the concept of DIPS. Voros McCracken first introduced the concept of DIPS back in 2001. Simply stated, DIPS maintains that pitchers don’t have any control over whether someone who puts the ball in play gets a hit off them. Whether a ball in play becomes a hit is determined solely by the ability of the hitter and the pitcher’s defense. If this concept is new to you, it can be difficult to swallow, but after causing a great deal of skepticism when it was first published, plenty of research has gone into the concept and with a few minor exceptions, it is generally accepted to be true.
So, then, why are some pitchers clearly better than others? Well, because pitchers do control certain aspects of the game. They control how many guys they walk; they can eliminate opponents with strikeouts; and they have control over how many homeruns they allow. K/9, BB/9, and HR/9 were the original cornerstones of DIPS, but when we are trying to evaluate how good a certain pitcher will be in the future we can look at a number of other aspects of his game.
Rather than continuing his work, McCracken went on to a front office job with the Red Sox, but DIPS and the proponents of it often make it seem like luck is the only factor to affect a pitcher other than the aforementioned rate stats. Team defenses vary greatly across the league, and the ability of a particular defense has a tremendous affect on pitchers. Groundball rates affect the HR/9 rate and also interact with the way a defense performs. Some pitchers can induce pop-ups as well and then there are always park factors to take into account.
For today, we will just focus on K/9 and BB/9. It is one thing to know that these rates affect a pitcher, but quite another to determine how much. By doing an extensive study on starting pitchers last year and determining how much their ERAs varied based on these rate stats I was able to calculate precisely how much these rates affected pitchers in 2006. Each K/9 affects a pitcher’s ERA by 0.14. Of the pitchers in my study, there was an average K rate of 6.58 and an average ERA of 4.52. That being the case, a pitcher who is average in every other way and gets 9 strikeouts per nine innings would be expected to have an ERA of 4.18, while a pitcher with a K/9 of 4 would have an ERA of 4.88.
Meanwhile, the BB/9 rate, which is often overlooked has an even greater effect on pitchers. Each walk per nine innings has an effect of 0.323 on the ERA. So with an average walk rate of 3.296, an average pitcher who walks 4 per nine innings should have an ERA of 4.75 while a pitcher who walks 2 would get a 4.10.
While these ERAs may seem high remember that we are talking about a league average pitcher who has league average defense. When you put the rates together, you can get somewhat more profound results. Johan Santana had a strikeout rate of 9.44 in 2006 and a walk rate of 1.81. A pitcher with these rates would be expected to have an ERA of 3.64.
That’s still too high! Well, yes, but remember that we are still looking at Santana as an average pitcher in every other way. In my next article I will examine homeruns and groundballs to see how much those affect a pitcher.