Statistical Prospect Evaluation Primer: Part 3: Pitchers

Part 3 in a series. You can read Part 1 here, and Part 2 here.

Pitchers

In the last part, I mentioned that I don't spend too much time looking at a hitter's triple slash line, but with pitchers I take an even more extreme approach; I almost completely ignore ERA. To understand why, you have to fast forward to what we know about pitching at the major league level.

I like to say that you need to do two of three things well to succeed as a major league pitcher; throw strikes (limit walks), miss bats (rack up the strikeouts), and keep the ball in the park (limit homeruns). Ideally, you'd like to do all three, but to have success at the major league level you only need to be able to do two of them at the same time.

As I explained in the Building a Better Projection piece, for the purposes of projection we rely heavily on Fielding Independent Pitching or pitcher-controlled stats. And since what we're doing when evaluating pitching prospects is projecting their future major league performance that's where we'll concentrate here as well.

Control

You can't teach power to a hitter or velocity to a pitcher, but perhaps the most teachable tool in all of baseball is the ability to put the ball in the strike zone. The Twins have built a pipeline of fine, young pitching, which allows them to compete on a medium-sized payroll, thanks to their strike-throwing philosophy. While not as proficient as the Twins, the White Sox, with most of the credit given to Don Cooper, have taught strike throwing at the major league level. Bobby Jenks, Matt Thornton, John Danks, and Gavin Floyd have all lowered their walk-rates while with the major league club.

When I spoke about walk rates in the hitting portion, I advocated the use of BB% or BB/PA, and the truth is that's the most accurate measurement for pitchers as well. But I still use BB/9IP. I'm almost ashamed to admit it. I understand that because of the variable, elastic length of an inning that any measure which relies on innings as a unit of measure will be inherently different from pitcher-to-pitcher. But I don't care. I've used BB/9 forever, and it's served me well.

As a general rule, I don't like to see a prospect have a BB/9 over 3, but I don't pay too much mind to walk-rate in the low minors. I'm looking for a progression in a pitcher's ability to throw strikes, or at least no regression as they face more advanced hitters. Major league hitters are going to spoil more pitcher's pitches, extending at-bats and coaxing more walks than their minor league counterparts, so that minor league BB/9 of 3 can rise to 4.5 in the majors in a heartbeat.

Strikeouts

Since control can be taught, when you're evaluating a pitching prospect, it's all about strikeouts. Again I'll use the more accessible K/9, which I can figure it in my head, over the more accurate K%, but I'll make no more apologies.

The magic number I'm looking for is 1 strikeout per inning or a K/9 of 9 or above. When you get above 9 K/9, that's where you find your aces. Tim Lincecum and Cole Hamels had minor league K/9 marks over 12. It wasn't exactly difficult to project future success for those two. It can be much more difficult for prospects who have K/9 marks at or below 9.

K/BB ratio -- I said I'd make no more apologies for using K/9 and BB/9, and here's why. K/BB is once again the best indicator of future major league success, and it doesn't have that pesky elastic-innings problem either. The magic ratio I'm looking for here is 3:1 or above.

Want to know why the Twins have been able to build a pipeline of young pitching? Look no further than the minor league K/BB numbers of their young starters.

• Scott Baker -- 4.25
• Francisco Liriano -- 3.25
• Nick Blackburn -- only 2.85, but 4.27 in AAA
• Kevin Slowey -- 6.94
• Glen Perkins -- 2.66 (Think there might be a reason he's had trouble staking claim to a rotation spot?)
• Boof Bonser -- 2.3 (same as above. Can't get his BB/9 under 3)

A prospect who keeps his strikeouts high and his walks low will find his way to the big leagues. But a prospect with a K/9 below 9 can succeed at the major league level, witness most White Sox pitching; they'll just need to be proficient in other areas, such as...

Groundball rate

When we talk about groundball rates, what we're really talking about is homers. I don't like to pay attention to minor league homerun rates at all for reasons that should become clear in this section.

Using Gavin Floyd as an example, in his last two seasons in AAA (spanning 220 innings) Floyd allowed a total of 18 homers. That's a pretty good total, especially considering he spent a good deal of that time in Charlotte's tiny ballpark. So it should have been a surprise when Floyd gave up 30 in his breakout season, right?

Wrong. Studies have found that major league pitchers on average give up a HR on about 12% of their flyballs allowed. I believe that number is lower in the minors (since they hit fewer HR), but I don't really care to know the specifics. The point is, if you want to know if a pitcher is likely to be homer-prone at the major leagues, look at his groundball rates.

There are a number of different measures for groundballs, and this is where it can get confusing. MiLB.com uses GO/FO, which is a measure of groundouts vs. flyouts, but that doesn't account for all balls in play, only outs. ESPN appears to have lumped balls categorized as line drives in with flyballs, which is how that recent THT article came to some of the wrong conclusions about Danks. Here I like to use GB% (found at firstinning.com) since it's the most accurate measure of a pitcher's groundball tendencies.

GB% is not without its pratfalls, however. First, the groundball rate in the minor leagues is higher than that in the majors. (Again, I'm not sure of the exact measure. But there is small, but noticeable difference.) And because there are 3 batted ball types (GroundBalls, FlyBalls, and Line Drives), 50% isn't the dividing line between a groundball or flyball pitcher. I would put that mark at a GB rate of about 45%.

What I look for, what the White Sox are looking for is guys who get groundball rates in excess of 50%. Coming back to Floyd, he had a groundball rate of 43% in each of his last two seasons in AAA, which explains the 47 homers he's allowed in under 300 innings for the Sox. Meanwhile, Jeff Marquez' groundball rates have been around 55% in his non-rehab assignments, while Jack Egbert has been consistently between 50-55%. The groundballingest of pitchers in the major leagues (Brandon Webb, Derek Lowe, et al) have GB% over 60%.

Conclusion?

If the hitting portion of this exercise could be summed up in two questions, the pitching portion requires three. Does he throw enough strikes? Does he miss enough bats? And What happens when hitters make contact?

By now you're probably going back through this list and thinking to yourself; doesn't Egbert fit these guidlines more than Danks? The answer is yes. So what's the reason Danks is the best pitcher on a major league club while Egbert has yet to get a taste of big league action?

There are two simple answers; Age vs. League, Danks was very young at each level while Egbert is over our guideline ages, and velocity, Danks is a lefty who sits in the low 90's with his fastball whereas Egbert is a right-hander who only touches 90-91 with his stuff. There is one more key difference that I'll illustrate in my final post on this subject.

Up next: wrap-up, and a look at a couple of current Sox major leaguers.