SwStk% and K/9
The SwStk% metric is one of my favorites on FanGraphs. We would expect the statistic to be a good proxy for a pitcher’s raw stuff, as a swinging strike is the ultimate pitch result. Therefore, it should correlate highly with strikeout rate. And intuitively, it does. From 2001-2010 using a sample size of 59 pitchers, SwStk% had a 0.87 correlation with K/9. It is not higher because luck still plays a role, and K/9 includes called strikes which obviously SwStk% does not. So now that we know SwStk% matches up with K/9 quite well, I like to rank pitchers by each to determine who might have some hidden strikeout rate upside or downside.
I looked at 92 pitchers from 2010 and compared how they ranked in SwStk% and K/9. A better ranking in K/9 than SwStk% indicates potential downside, while a better ranking in SwStk% would suggest potential upside.
First, I present the 10 pitchers with the largest difference in the two metrics where the ranking in SwStk% is better than K/9. These are the upside guys.
Clay Buchholz stands out near the top of this table. While the majority of fantasy owners rightfully expected some major ERA regression in 2011, I personally did not believe that regression would lead as high as his 2010 xFIP of 4.07. The scouting of his stuff supports his strong SwStk% so I think he has significant K/9 upside, which should offset some of the luck neutralization this year. Hiroki Kuroda‘s SwStrk% has never been below 9.2, yet his K/9 has never matched up with these strong marks. Kuroda is always undervalued in fantasy leagues to begin with so the possible promise of a higher strikeout rate is even more of a reason to get excited. If only Jaime Garcia’s innings jump did not scare me away, he would be one of the more intriguing pitchers for this season, given his strikeout and ground ball combination. Shaun Marcum could be a darkhorse Cy Young candidate after his move into the National League, but his past injury issues, including this spring training, caused me to shy away. If Johnny Cueto can return fully healthy and his increased ground ball rate in the second half of last season is sustained, he could actually post another mid-3.00 ERA again, but this time fully supported by his skills.
And now for the downside guys.
That is a huge difference for Yovani Gallardo. He has never posted SwStk% marks that would match his ERA, as his career best is a solid, yet unspectacular 9.2%. With only a league average ground ball rate and mediocre control, he is a surprise bust candidate. C.J. Wilson owners already have to worry about his huge innings jump of last year, but now the prospect of a decreased strikeout rate should have owners fearing the worst. I would not be surprised if Wilson is worthless in 12-team mixed leagues. Those of us, including me, who were optimistic about Jason Hammel’s breakout chances should be a little less so after seeing his name above. Add another mark against Ubaldo Jimenez for his chances of coming anywhere close to a repeat of 2010 or just breaking even for his fantasy owners who drafted him this season. Tommy Hanson continues to get hyped as one of the best pitchers in the game, but I struggle to see why. Admittedly, his stuff appears top-notch, but it simply has not generated results as strong as should be expected. His second half last year included a major K/9 decline to just a 6.4 mark and with just three strikeouts in his first nine innings to open the season, combined with a fastball velocity drop of 1.6 miles per hour, owners should be very concerned.
As a reminder, this analysis is not perfect. I am completely ignoring called strikes, which is a skill, and is a factor that influences strikeout rate of course. Some of the pitchers on the above lists may have called strike rates low or high enough that would indicate that his strikeout rate is not actually in for a fall or increase.
Mike Podhorzer is the founder of ProjectingX IQ, an advanced fantasy baseball analytics platform that transforms projection data and in-season performance signals into actionable intelligence. He is the 2015 Fantasy Sports Writers Association Baseball Writer of the Year and three-time Tout Wars champion. He is the author of the eBook Projecting X 2.0: How to Forecast Baseball Player Performance, which teaches you how to project players yourself. Follow Mike on X@MikePodhorzer and contact him via email.
Genuis!
I’m an idiot and spelled the single word in my post wrong, but you get the idea of how I feel about this article…a great way to identify undervalued pitchers or those ready to take the next step
Called Strike Rates do NOT have really any impact at all on a pitcher’s K rate. The correlation is basically 0.
As for what besides swinging strike rates causes the increase in Ks? Well, I’d say luck is a very relevant factor.
I think foul rate of pitches should explain a good portion of K/9 rate but i am too busy to analyze it (cramming for exams
Question: Does research show that the relationship between swinging strike ratio and K/9 regresses towards the mean? Or is it possible that the differences are caused by differing pitcher styles? I think that’s probably what’s happening with Dickey and possibly some of the other guys.
Very interesting, but I don’t think we have any idea what it all means. Just looking at the names on the “upside” and “downside” lists, it does not appear that there is any great trend toward inconsistency on either side.
so is Gallardo just the K/9 lotto winner? I noticed his swinging strike rate in the offseason, but I figured 500 IP of about a K/IP would show that he’s doing something to “earn” those Ks. I do agree though, that if the K-rate dips to the low 8s, he’s probably roughly a 4 ERA/1.3+ WHIP pitcher barring improved control.
This is a well written piece that is tremendously informative and interesting. Thank you.
I know I busted your nuts the other day, but this is good stuff. A few things that I’d suggest doing while you have the data open: Check the relationship between SwStr% and not K/9, but K:AB or K:PA K-Rate, whatever. Just anything that quantifies strike-out ability on a per batter rather than per inning basis.
Secondly, just eying your data — please remember I’m throwing darts blind-folded and drunk at the board here — I think that you might find an interesting correlation in K/9 and optimum pitch or top-2 pitch SwStr%.
I’d posit that someone could have a nice high swinging strike rate if they featured multiple above average pitches (curve+change based on that list) and fastball movement but suffer in K/9 without a true *out pitch* that has a well above average SwStr%.
It would be interesting (and you have the data open, right!) to see this very same methodology repeated with SwStr% on a pitcher’s best and then best and second best pitch. I bet the top two would give you a very nice corr. to K-Rate.
That top list just really reminds me of a list of dudes that throw their fastball less than 55% of the time.. right? Anyways. Goodstuff!
This strikes me as a really smart possible explanation. If a pitcher is throwing fastballs to get ahead in the count and getting them fouled off but has a nasty change that he’ll use to K up the hitter, then he’ll still have a good K/9 despite the lower SwStr%. In fact, this could be preferable.
Really nice post, but I would like to see a formula that predicts k/9 based on SwStr% instead of rankings. Using rankings doesn’t actually show the magnitude of the gap between their k/9 and their predicted k/9… even with a rather large ranking gap, it may be a trivial difference since I’d expect all pitchers to be relatively closely packed together.
There was a formula published in an article in the last year, but unfortunately I haven’t been able to find it again!
I think jaime garcia’s swinging strike rate is finally showing up in his K rate… 18 Ks in 15 innings. I was watching today and it felt like every single PA had at least one swinging strike at a live two-seamer or a changeup dropping off the plate. It was an absurd amount of swings and misses. I wish he would have stayed in the game longer than 6 innings when he is so clearly dominating and only at about 80 pitches or so, but that’s the National League for ya…
Very interesting! The good news: I am a Garcia owner an one league. The bad news: In another league, I own Cliff Lee, Ian Kennedy AND Jason Hammel. Ouch. I haven’t looked at Lee’s looking-strike rate, but I’d imagine it’s pretty solid.
kuroda is sick, hammels is ok. cliff lee didnt get help from victorino (bad OF play that caused a few runs) defense is key too. gallardo gets great defense from his guys. dont pick any marlin or wash. pitchers, however wash now has desmond/esponsoia/werth. STL pitchers are not getting any run support or D, franklin is a joke closer too (see rasmus basket catch?). Kuroda has been sick so far and did not need any run support. broxton is a little concerning though.
dont get any cub pitchers, their defense sucks. best of luck
Gallardo gets great defense?
The dispersion of swinging strike % is tight.
In 2010, among 151 pitchers who threw >140 innings, 105 had a swinging strike % between 6.7-9.7%… of course, that means that in a 100 pitch outing, the difference between the 23rd-and 108th-best swinging strike pitchers in the game is three swings and misses. Three swings and misses! Three pitches!
While those three picthes are probably worth, say an extra 1-1.5 K/outing, it makes sense that swinging strike % correlates with K/9. But I would not extrapolate any greater signifncance from the data than that… and certainly would not suggest that the effectiveness of these 3 pitches per 100 are a harbinger of regression or progression. I would suggest that the location of the 90 pitches that are not swung and missed at is much more important.
Take the cluster of pitchers who averaged 6.7-6.8% swings and misses in 2010 (SwStr% ranks 103-108): Randy Wolf (100, 215), CJ Wilson (76, 204), Bronson Arroyo (93, 215), Tim Hudson (72, 229), Fausto Carmona (88, 210)and Armando Gallaraga (72, 144). (ER, IP in parentheses) totals: 501 ER, 1217 IP = 3.71 ERA
And lets compare them to the cluster of pitchers averaging 10.7-10.9% swings and misses (SwStr ranks 23-26): Bud Norris (84, 154), Shaun Marcum (79, 195), Ryan Dempster (92, 215) and Huroki Kuroda (74, 196). totals: 329 ER, 760 IP = 3.89 ERA.
So although swinging strike % intuitively will lead to a higher strikeout rate (and perhaps an extra 30-40 strikeouts per season, 1-2 strikeouts per week) for a pitcher, I caution against overinterpreting your data to suggest that Yovani Gallardo “is a suprise bust candidate” or that Shaun Marcum is a “darkhorse Cy Young candidate.” You are making these statements based on three pitches per outing for each of these pitchers.
P.S. Please forgive me if the math is a little off. I am making these calculations on a napkin during my lunch break and without a calculator.
based on these small samples, within a standard deviation of the mean, there seems to be little correlation between SwStr% and ERA.
does anyone have data on Swinging Strike rate by pitch count?
or just all 0, 1, and 2 strike counts?
You have a couple choices: Build PFX database or beg (with a donation, some may call it a bribe) Trip over at Texasleaguers.com to spit out the query for you.
Building your own PFX database is FUN and EXCITING and TOTALLY NOT TIRESOME AND BORING.
Last year, I took a look at the relation between K/9 and SwStr% and how changes in latter affect the former. Needless to say they have a very high correlation. More interesting, however, might be to compare K% to K/9
http://www.hardballtimes.com/main/fantasy/article/the-sum-of-their-parts-a-deeper-look-at-peripheral-data/
Funny, I was the first to leave a comment on that article and suggested using K%! I wish that and BB% was available on here.
Touche sir. I’ve seen become a K% fan.
Also, all you need to do is Divide BB/TBF or K/TBF.
I had book marked this article…. and wanted to check out how your thoughts were doing.
So far the Marcum v Yago is spot on.
How does SwStk% advance our knowledge of a pitcher’s skill over K/9? Isn’t the ultimate result the “K”–whether swinging or taking? If Gallardo gets more strike calls without a swing, is that less valuable? Not sure what we are learning here.
In my opinion it increases our knowledge into K/9 and can be an indicator on whether or not a pitchers K/9 can increase going forward. Yes there are pitchers that hit the corner and are control pitchers that might have a lower swinging strike rate compared to their K/9 but i think it is a beneficial stat none-the-less.