Minor League K/9 And Velocity
Sorry for the absence of posts yesterday – I was pretty under the weather.
Getting back to the charting from a few days ago, but with a little bit of a twist. This next graph is similar to the previous ones, plotting fastball velocity and strikeout rate, but in this case, it’s minor league K/9. I took all pitchers who threw at least 30 innings in the minors last year and at least 10 innings in the majors and tied the major league velocities (published here on the site) to their minor league K/9 rates.
Essentially, the goal is to see whether there is a bigger variance in K/9 from velocity in the minors, which is what we believe to be true based on anecdotal evidence. We’ve all seen plenty of junkballers with 85 MPH fastballs blow away Double-A and Triple-A hitters with command and movement, but not be able to repeat this success in the majors. If our anecdotal evidence is correct, and there are significant amounts of minor league pitchers who can post miss bats with questionable fastballs, they should hopefully show up on this chart.
Obviously, there are selection bias issues here, as we only have velocity data for guys who were given some time in the majors. Pitchers who don’t get a callup to the big leagues won’t appear in our data set. It’s a problem, but we deal with what we have. Anyway, here’s the graph.
First thing you should notice is that the starting point on the regression line is quite a bit higher than in the major league graphs. The really low strikeout minor leaguers just don’t get called up. The lowest K/9 in the sample of 183 pitchers is Brad Hennessey, who posted a 4.69 K/9 for Fresno. His fastball was 88.3 MPH in the majors.
The other thing to notice is that the trendline is much flatter, and while the r of .27 means that the spread isn’t totally random, there are a ton of data points that don’t fit the line. Clay Rapada, for instance – 86.5 MPH fastball, 11.57 K/9 for Toledo, or David Robertson’s 13.11 K/9 with a 90.8 MPH fastball. Jason Bulger had the highest strikeout rate in the sample at 15.7 K/9, and while he’s not a soft-tosser, his 92.9 MPH fastball isn’t that far from the median.
While this isn’t a perfect analysis, it does seem to confirm the common wisdom that velocity isn’t as necessary to rack up strikeouts at the minor league level. There are a good number of pitchers who miss bats without a big fastball. When looking at minor league pitching numbers, you have to keep this in mind – gaudy statistics do not mean that you can automatically infer major league quality stuff.
Print This Post
Using the past three years and all pitchers with over 200 IP I plotted K/9 vs. O-Swing % and found a correlation almost as strong as K/9 vs. FB velocity. What was really interesting was that there was very little correlation between FB velocity and O-Swing %. The pitchers who post K/9 rates greater than their velocity would dictate do so by getting hitters to swing at pitches outside of the zone.
Also there was no correlation between K/9 and % of pitches thrown in the strike zone or between K/9 and first pitch strike % (although both of these correlate strongly with BB/9).
you should control for age in your analysis. i would think it is a confounder that probably attenuates the relationship.
Someone has got to come up with a way to account for a great secondary pitch for relievers. Especially for a loogy type like Rapada, the 85 mph fastball doesn’t hold him back because against most lefties, they’re seeing probably 50% sliders, and he’s got a good one. It seems like with specialized bullpen roles, there are ways to get strikeouts without a good fastball.
I’m guessing Jason Motte is that speck on the high and far right side.
I think it is fair to say that “gaudy statistics do not mean that you can automatically infer major league quality stuff.” But can you infer the opposite? meaning that a lack of gaudy k-rate statistics in the minors means you can automatically infer a lack of major league quality stuff”? I suppose there are exceptions that prove this rule, like rick porcello. But generally, I think of minor league K/9 like this: If you have a good/great rate, it doesn’t mean you will be great in the majors. But if you don’t have a good/great rate in the minors, you probably don’t have major league quality stuff. I look at mostly like a high minor k/9 rate is more of a necessary rather than sufficient condition.
What I think would be interesting to see is a comparison of slow throwers minor league to major league k/9, and a comparison of hard throwers major league to minor league k/9. So maybe a graph of the difference in majorl eague and minor league k/9 vs. velocity. That would get rid of the selection bias issues, and I think would answer the question about soft throwers being able to strike out minor leaguers and not major leaguers more effectively.
Is this blog entry missing a graph? From the text I expect to see one graph for minor league data and one graph for major league data.
There was a post the other day where he compared velocity to major league k/9, today he’s comparing velocity to minor league k/9.
I think Dan hit the nail on the head–there must be adjustments for age (probably handedness, and role [starters v. relievers] as well).