What Can We Learn From Expected Strikeout and Walk Rate Outliers?

Sabathia is telling his secret to having fewer strikeouts than expected. - Nick Laham

Which pitchers strikeout and walk more or fewer batters than we expect and what similarities do they have?

Introduction

Two weeks ago, Glen Perkins stated in a Fangraphs Q&A: "I don’t know if there’s a stat for an expected strikeout rate based on swings-and-misses and chases." Well obviously, I went into overdrive attempting to create such a stat. However, my quick work was in vain, as Beyond the Box Score's very own Blake Murphy did exactly that in his series on predicting walk and strikeout rates the day before the Perkins Q&A was posted.

Since Blake was ahead of us all on this front, I want to attack the idea from a slightly different perspective: which pitchers consistently outperform our expectations based on their peripherals and what can we learn from them?

Method

Just to keep things interesting, I ran my own multiple regression to find expected strikeout and unintentional walk and rates. I used the same minimum of 350 total batters faced to keep my method similar to Blake's.

Expected K/TBF = 0.02-Swing%(pfx)*0.71+Zone%(pfx)*0.37+F-Strike%*0.12+O-Swing%(pfx)*0.15+SwStr%*2.3

Adjusted R squared 0.75, Standard Error 0.02

Blake used only SwStr%, fastball velocity, and league average to create his equation, with basically the same results as mine. I missed fastball velocity when initially running the calculations and feel like his simpler and as accurate equation is a much better route.

Expected UIBB/TBF = 0.37-O-Swing%(pfx)*0.32-Zone%(pfx)*0.12-F-Strike%*0.23-Z-Swing(pfx)*0.07+SwStr%*0.41

Adjusted R squared 0.57, Standard Error 0.01

Blake did a similar shotgun technique to me with similar results. It's harder to predict walk rate based on peripheral stats than it is for strikeouts. I wonder if this is because there is not as much variability in walks as in strikeouts.

If we look at seasonal K/PA, the mean is 0.17, with a range from 0.05 to 0.30 and a standard deviation of 0.04. Compare that to UIBB/PA: Mean 0.07, range 0.02 to 0.17, standard deviation of only 0.02. Walk rates are much more bunched together than strikeout rates.

There are low strikeout pitchers and high strikeout pitchers, but you can't make it in the bigs if you walk a lot of people. The top two pitchers in UIBB/PA since 2007 are Dennis Sarfate (23 innings pitched after a 0.17 BB rate in 2008) and Kyle Drabek (71 innings pitched after a 0.15 BB rate in 2011).

Based on these equations, which pitchers consistently outperform their expected rates?

I took the expected rate and multiplied by the total batters faced that season, then subtracted this from the actual amount of strikeouts and walks. So in 2010, Cliff Lee had 185 Ks and 16 UIBBs, but was expected to have 159 and 25 based on his peripherals, for differences of 26 and negative 9, respectively.

I did this for each player season since 2007 and added up all the differences to find the top outliers.

Results

Strikeout Rate

Here are the leaders with at least 1,000 batters faced:

Player Extra Ks Per PA
Vance Worley 0.055
Yovani Gallardo 0.048
Erik Bedard 0.039
Travis Wood 0.038
Brandon Morrow 0.037

What a weird group. Worley isn't exactly a strikeout pitcher, hovering just above league average. In the two full years he has appeared in the major leagues, he has a total of 226 strikeouts. However, based on his peripherals, we would have expected him to only rack up 163. Travis Wood is in a similar boat. Gallardo on the other hand, is definitely a strikeout pitcher, ranking 21st in the league since 2007. Bedard and Morrow also have a propensity to get swings and misses.

Of these top five, Wood is the only one who regularly throws a changeup. He's also the only one to heavily throw three different fastballs (four seam, sinker, cutter).

And pitchers with the most strikeouts missing per plate appearance based on their peripherals, minimum 1,000 batters faced:

Player Extra Ks Per PA
Scott Olsen -0.048
Tim Wakefield -0.039
Jeff Karstens -0.035
Paul Byrd -0.034
Chris Capuano -0.032

Okay, so a knuckle-baller broke the system. Beyond that, we have two lefties and two righties. What do Olsen, Karstens, Capuano, and Byrd (and Wakefield, for that matter) have in common, according to Brooks Baseball?

Player vFA
Scott Olsen 89.06
Chris Capuano 88.45
Paul Byrd 86.6
Jeff Karstens 89.97

They're all soft-tossers. So I really should have included fastball velocity in my metric like Blake did. Interestingly enough, all four of them also rely heavily on fastball-slider-changeup repertoire, though I'm not sure how common that is.

This chart summarizes the findings:

Pitcherskrate

Jonny Venters sticks out like a sore thumb on the bottom right, but a full 46% of the players with a strikeout rate above 0.20 and fewer strikeouts than expected per plate appearance are left handed pitchers. Compare this to 28% of all pitchers in this study who are left-handed and we have a clear trend. I wonder if some of this is due to lefties facing more right handed hitters. They are less likely to strike out the hitters that they see more often, leading to lower than expected strikeout rates. This isn't the case with every lefty though, as both Bedard and Wood are top-ranked at getting more strikeouts than we would expect.

Walk Rate

Player Extra BBs Per PA
Hiroki Kuroda -0.025
Jeff Francis -0.021
Brett Anderson -0.020
Livan Hernandez -0.020
Jesse Litsch -0.020

What do the top five pitchers who outperform their peripherals and walk fewer people than we expect have in common? Not throwing the four seam fastball a lot.

Kuroda, Francis and Hernandez are all sinkerballers, throwing their two seamer a majority of the time. Anderson does throw his four seamer more often than any other pitch, but only does so at a 41% clip. He throws a slider nearly as often at a 33% rate. Litsch throws his four seamer, two seamer and slider at comparable rates, but throws a cutter more often than either of those three.

Now the pitchers who walk more than we expect:

Player Extra BBs Per PA
Chris Young 0.028
Daisuke Matsuzaka 0.026
Jonathan Sanchez 0.024
Micah Owings 0.020
Oliver Perez 0.020

Sanchez and Perez are the only two of these five to appear in the major leagues this season. It's not going very well for Sanchez, who has a 9.68 FIP in 13 innings, although his walk rate has fallen to a more manageable 0.11. Perez is a little better off with a 3.67 FIP.

Pitchersbbrate

Pitchers that walk a bunch of people tend to walk even more hitters than we would expect. It's a linear trend and is much more clear for walks than for strikeouts. Perhaps this is why walk rate is a better next season predictor than anything else. I had a classmate in college that would get into failure spirals. Once one thing went wrong, he would get angry, leading to another thing going wrong, leading to more anger. It was rather funny to watch from a distance, providing he did no real harm to himself or others. I feel like that's what happens with pitchers like Young and Matsuzaka. Once the walks start, they just keep coming and coming until eventually they've lost control of the game.

Summary

So what have we learned? Based on the preceding small sample sizes, this may be worth more digging into pitch usage data.

Does not having a changeup help pitchers get more strikeouts than expected? Do pitchers with fewer than expected walks rely less on their four seam fastball?

And a few things are certain: Lower fastball velocity means fewer strikeouts, left handed pitchers are more likely to strike out fewer batters than we expect, and finally high-walk pitchers just keep right on walking people.

As usual, thanks to Fangraphs for the data.

Btbs-twitter-insert_medium

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Beyond the Box Score

You must be a member of Beyond the Box Score to participate.

We have our own Community Guidelines at Beyond the Box Score. You should read them.

Join Beyond the Box Score

You must be a member of Beyond the Box Score to participate.

We have our own Community Guidelines at Beyond the Box Score. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9351_tracker