The Hardest Fastball to Hit?
I am obsessed with evaluating pitchers and pitches. As part of a data request from Jeff Sullivan on the percentage of swinging strikes that pitchers generated off fastballs last year, I went looking through the PITCH f/x data and ultimately ended up generating a chart grouping each different pitch type from every pitcher in the big leagues in 2008, separated by whether or not the pitch was thrown to the same or opposite-handed hitter and whether the pitcher was starting or relieving.
That is, I have the results (in terms of result of the pitch) and can compare the results of, for example, fastballs thrown by Joba Chamberlain:
As a starter against a right-handed batter.
As a starter against a left-handed or switch-hitting batter.
As a reliever against a right-handed batter.
As a reliever against a left-handed or switch-hitting batter.
And so on, across each different pitch that he, or anyone, throws. This dataset is going to form the basis for more than a few posts going forward*, starting with this one. Among the first things that I did was strip out all instances of pitch totals under 100, feeling that 100 is a pretty decent sample of pitches given my strict categorization. Sorting that data by the percentage of pitches swung on and missed, I encountered right at the top, an amazing number. Ryan Madson’s changeup (as classified by MLBAM), thrown to same-handed hitters generated a swing and miss a whopping 36% of the time, about 5% higher than any other pitch by any other pitcher in 2008.
Scrolling down the list, I also noticed (not to my surprise) a lack of fastballs showing up. So I decided to see which pitcher first popped up. That name turned out to be Brandon Morrow. His fastballs as a reliever to right-handed batters clocked in at an impressive 20% swinging strike rate. What’s curious though was the spread in his results across the four categories. Here they are, presented in the same order listed above.
SP, Same hand – 10% (135 pitches)
SP, Opp hand – 7% (183 pitches)
RP, Same hand – 20% (219 pitches)
RP, Opp hand – 9% (216 pitches)
The 20% figure suck sticks out, doesn’t it? It’s worth pointing out that the average rate for fastballs from starting pitchers last year was just under 6%, so even at his worst, Morrow’s fastball appears to be an above average weapon for him. Still, it will be interesting to begin contrasting and adding 2009 data to this when the season begins.
*If there’s a question you want answered that you think this data can help with, shoot me an e-mail and I might be able to look into it for you.
Matthew Carruth is a software engineer who has been fascinated with baseball statistics since age five. When not dissecting baseball, he is watching hockey or playing soccer.
Do we have any information on when the sample size becomes significant for something like this? I’d love for this to be a significant sample as it bodes great for him as a starter (and as a closer as well I suppose, but hopefully that shouldn’t ever be relevant again). 753 pitches to 265 total batters, any input from the fancy statisticians around here? I would certainly think it indicates it’s a quality weapon against lefties, but how much can we expect it to reasonably regress?
26%
That’s alot.
Matthew,
I’d love to know the overall composite distribution of results per pitch for starters. Your, 6% swinging strike rate for fastballs from starters is an interesting figure – I would have guessed slightly higher.
So.. what about the rate of taken strike, foul, ball and ball in play? (maybe need an other for hbp…anyway) And then how does this distribution compare to other common pitches like curves, sliders and change-ups? Am I blowing the cover of another blog post?
This is great info. Gotta love Pitch F/X data.
Nice post. I’d be interested to see the average increase in the swinging strike rate when a pitcher moves to the bullpen from the rotation. I know sample size may be an issue, as not many guys get significant time as both a starter and a reliever in one season, but I think it would be a worthwhile look.
I’ve seen a lot of the pitch f/x data analyzed in a very cool way lately. Thanks for that!
However, I do have one major concern about all this: none of these data (i.e. swinging strikes) exist in a vacuum. If hitting is timing, and pitching is all about upsetting timing, then I think it would be a worthwhile exercise to examine pitch sequencing within a pitcher/ hitter encounter.
For Brandon Morrow, it’d be _really_ interesting to see something like “for his relief appearances, when he threw a fastball preceeded by a change, he saw x% increase in the likelihood of a whiff on the fastball…”
To me, that would be more interesting, because those data could be prescriptive — it could affect pitch sequencing in games. I would be interested in working on something like this, if you guys are willing to share the data that you’ve collected and processed.
Thanks!
Travis
Thanks a lot for the information, it’s always interesting to see stats given in this way. I agree it would be perhaps more interesting to see more prescriptive data.
Thanks.
At last, I found the right place to read some quite interesting stuff, thanks for sharing.
Hi, I really beautiful and useful content sharing … Thanks
That’s Wonderful Thanks…
hello !
Well, I am so excited that I have located your post because I have been searching for some information on this for almost three hours! You’ve helped me a lot indeed and by scaning this post I have found many new and useful info about this subject!
porno adult sitemi eklerseniz sevinirim
Thanks for all 🙂