2023 Outliers: SwStr% vs. K%
I look into the relationship between SwStr% and K%, and find SP outliers on both sides of the trend line
One of my personal favorite things to do in the baseball analytics world is to find two stats that are highly correlated and then look for outliers in the data. This works best when we have an “input” and an “output” stat, meaning that one stat is more granular and does well to predict the other stat.
What we’re looking at here is swinging-strike rate (SwStr% = Whiffs/Pitches Thrown) and strikeout rate (K% = Strikeouts / Batters Faced). We all know that it’s incredibly important for a pitcher to have a high strikeout rate if they’re going to have success. I don’t think I need to prove that to anybody. The best “advanced” stat for predicting K% that I have found is SwStr%, so I really like to look at SwStr% often.
In this short post, I will talk briefly about the relationship between these two stats, give you data and plots, and then point out the biggest outliers on both sides of the trend line here.
The Relationship
The quickest and most visually appealing way to learn about statistical relationships is with a scatter plot. I looked at 2021-2023 (taken as individual seasons) and found swinging-strike rates
We see a clear linear relationship here, meaning that as one number goes up - the other tends to go in the same direction at the same relative speed.
To put numbers on these:
Correlation coefficient: 0.86
R-Squared value: 0.74
I’ll spare you the statistics talk, but just know that anything close to one in those values represents a very, very strong positive relationship. Since we’re over 0.7 in both - it’s a very steady and reliable relationship to study. Stated plainly:
we should not expect to find a low K% with a high SwStr%, and vice-versa.
The Expectation
Here’s a quick “mapping” table of what general K% should be expected with every SwStr%:
SwStr% → K%
8% SwStr% → 15% K%
9% SwStr% → 17% K%
10% SwStr% → 18% K%
11% SwStr% → 21% K%
12% SwStr% → 22% K%
13% SwStr% → 24% K%
14% SwStr% → 26% K%
15% SwStr% → 27% K%
16% SwStr% → 30% K%
17% SwStr% → 33% K%
There are very few pitchers that are able to exceed 16% SwStr%. The only pitchers to do that in the last three years (given at least 10 starts):
Rodon, Kershaw, Burnes, Cease, deGrom, Gausman, Giolito, Scherzer, Ray, Bieber, McClanahan, Glasnow, Heaney, Cole, Ohtani, Strider, Snell, Eury Perez, Peralta, Skubal
And it’s pretty tough to stay in the big leagues with a SwStr% under 8%. Those names:
Brett Anderson, Arrieta, Arihara, Wainwright, Houser, Senzatela, Adon, Jason Alexander, Fedde, Dakota Hudson, Winckowski, Jake Irvin, Martin Perez, Ty Blach
The 2023 Outliers
So to find the outliers, we’re looking at the player dots that are furthest away from the trend line. You can also take the percentile for each and subtract them to find the biggest differences.
Negative Outliers
Players who over-performed in K% according to their SwStr%
Hunter Brown…….11.5% SwStr% → 26.7% K%
Logan Webb………..9.9% SwStr% → 22.8% K%
Seth Lugo……………10.1% SwStr% → 23.2% K%
Mitch Keller…………11.1% SwStr% → 25.5% K%
Javier Assad………....8.2% SwStr% → 21.0% K%
Positive Outliers
Players who under-performed in K% according to their SwStr%
Jose Urquidy…………..12.9% SwStr% → 16.4% K%
Tyler Anderson………13.8% SwStr% → 18.9% K%
Sandy Alcantara……..13.7% SwStr% → 19.8% K%
Carlos Carrasco………11.6% SwStr% → 15.8% K%
Roansy Contreras…..12.9% SwStr% → 18.3% K%
You’re probably not surprised to see Logan Webb here, but really this is the first time he’s been a true outlier in this kind of study:
So the SwStr% has skipped around, but in 2021 and 2022, the K% lined up right where we expected with the input SwStr%. In 2023, however, the SwStr% went way down under 10% but the K% stayed above the league average at 23%.
The problem with him is that we know he can get whiffs if he really wants to, as given by that 13.4% SwStr% in 2021 - he just seems to be someone who is choosing to stick to just generating soft contact (something that not many people can do consistently - but he clearly can). It’s possible he will change approach next year and we see the SwStr% over 12% again, it’s hard to say - so I wouldn’t make any overly strong claims about him.
The other expected outlier here is Sandy Alcantara, who has long been posting high swinging-strike rates with underwhelming strikeout rates.
That said, the K% bottomed out last year, so yes I would absolutely expect a higher K% next year - EXCEPT remember that it doesn’t matter for us right now because he’s out for the whole 2024 season after undergoing elbow surgery this fall.
I will provide my favorite buy-low SPs for 2024 (according to this study), the link to the full plot and give you the full data below the paywall. Sign up today for $7/month to get everything I’m offering this offseason, which will include
Team-by-team 2024 previews (will come out rapidly starting in January)
My projection model’s 2024 full-season projections (new and improved, release date late December/early January)
More posts, plots, and data like this one