This week, we’ve been discussing Jon Niese’s fastball a lot. First, I compared Jon Niese’s fastball to other successful MLB lefties and concluded that he had enough velocity to be successful. Second, I examined the claim that Niese’s fastball was “too straight” and found little merit to the claim this year.
However, our resident pitch fx Josh Smolow went through the data to differentiate between Niese’s two and four-seam fastballs this year. Remember, Josh wrote about Niese earlier this season, and discussed the development of Niese’s two seamer, which he uses to induce groundballs.
So lets turn this over to Josh:
| Mean Speed | Mean Horizontal Movement | Mean Vertical Movement | |
|---|---|---|---|
| 2008 Fastball (3 starts) | 89.14 | 3.51! | 8.54 |
| 2009 4-Seam FB (2 starts) | 89.88 | 5.72 | 9.53 |
| 2009 2-seam FR (2 starts) | 89.7 | 10.5! | 6.24 |
These numbers are more of what you saw from Toby, but bear with me here, it gets better. The numbers that are of interest here are the Mean Horizontal Movements of his fastballs.
Essentially, you can see that Niese’s fastball was ridiculously straight last year (moved only 3.5 inches horizontally in on lefty batters). This is a composite of ALL fastballs thrown by Niese last year, so odds are that number is actually HIGH for Niese’s four seam fastball…which is a clear sign in the poor quality of that pitch last year. (Toby: Because the sample likely includes some poorly differentiated 2-seamers.)
As you can see from the numbers above, both the horizontal movement of the four seam fastball and two seamer have increased dramatically from last year.
Numbers are a bit hard to picture in your head…but a picture isn’t. Behold:
The image to the right represents a bird’s eye view of the average fastballs that Niese has thrown the last two years. As such, the paths represent the horizontal movement of the fastballs as they leave the pitcher’s hand (the top of the graph) and go toward the Plate ( the line at the bottom of the plate).
There are 3 paths…the blue path represents the path of the Niese’s average two seam fastball in 2009, the green path represents the path of Niese’s average four seam fastball in 2009, and the white path represents the general path of all of Niese’s fastballs last year (2008).
You can truly see the paths of the pitches on this chart and the movement in the pitches. If there was no movement at all on the pitches, the pitches would go in a straight line (diagonally)…but instead you can see each paths curve back somewhat toward the left handed batter. Here you can truly see the 4+ inch difference in the movement of the two seam fastball from the four seam fastball in 2009. The four seam’s movement would seam to be more predictable to a batter, whereas the two seam seems to truly bend and move further than you’d expect.
Then there’s the 2008 fastball. It’s kind of hard to see, since the pitch seems to start at a different release point than the 2009 pitches,* but you’ll notice it doesn’t bend really that much at all. The pitch is indeed, as people have noted, really “straight.”
*Before people comment, this release point difference is NOT significant of anything. PitchFX cameras tend to vary the initial release points of pitches per starts. There is no clear evidence Niese’s fastball release point has shifted.
One final thing you can see on this graph. Niese’s fastball reaches the batter in roughly 4 tenths of a second. Each dot on each path represents .025 seconds, so that four dots represents about ¼ of the time it takes for the pitch to reach the plate. You’ll notice that by the 8th dot, or when the pitch is midway between the pitcher and the batter, the four seam and two seam fastball seem to be exactly identical (in the horizontal plane at least).
A batter needs to decide whether to swing at a particular pitch around (roughly) that 8th dot. So a batter can’t tell whether a fastball is the two seamer or the four seamer before he has to decide whether or not to swing at the pitch. He has to guess.
Advantage: Niese.
awesome. really awesome.
but if you convince me to like niese and then he sucks the big one, i’m going to punch you.
excellent write-up, the picture is what makes it really great though…
What do you make of the two seamer (sinker) having less vertical movement than the four seamer? Is this unusually low and could he improve this pitch with more v-move?
Nate, no this is particularly good….Consider a two seam fastball to act more like a sinker than the four seam fastball. The difference between the two’s vertical movements means that the pitch appears to the batter to DROP and thus should theoretically result in more ground balls.
This extra drop is a positive development that Niese didn’t have last year…whereas last year he had a middle of the line fastball that didn’t “rise” or “drop”, this year he has a “rising” four seamer and “dropping” two seamer.
…the Mets should promote Niese so this data can be expanded upon…
well that and he would be their third best starter right now.
I got a question.
There was talk last season of Niese developing a cut fastball. If he used this cut fastball last year when he got called up might that be a reason for the low horizontal movement?
Last year I had done a pitch fx profile on him
http://www.metsgeek.com/articles/2008/12/16/jon-niese-season-analysis-2008/
It does indeed look like he had little horizontal movement on the fastball last season. One thing to keep in mind is the movement values vary greatly by ballpark and sometimes even by game in the same ballpark. Alot of work is currently being done to adjust those values. So we definitely need to take the movement values with a grain of salt.
Another thing here is we are talking about only a handful of starts per year. We need to be extremely careful on what we infer from it.
1. Agreed whole-heartedly on the small sample sizes. Any conclusions we draw here must be very tentative.
2. I’ll defer to Josh re: movement variations park-to-park.
3. On the cutter question, Niese and his pitching coach at AA, Ricky Bones each told me last summer that he was learning a cutter. I saw no evidence of that pitch in his big league outings in 2008. In 2009, I saw two pitches that look like his cutter in the pitch fx data that are mislabeled sliders. In his recent AAA starts, Niese was throwing his cutter, which looks like a hard slider, significantly more.
Hey John, i think i commented on your Mike Pelfrey article last year.
To answer your questions, there’s really no evidence of a “cut fastball” last year or this year….if he was using such a pitch, then he wasn’t actually using any other type of fastball this year or last (making it kind of pointless).
Sidenote: I mean, the Horizontal movement’s difference between his four seamer and two seamer this year is so much that if he threw each pitch equally as much the four seamer would focus as a cutter. But he primarily throws the four seamer here.
I’m aware of the handful of starts and the park effects on pitchfx (Harry Pavladis is working on that and i’m looking forward to seeing his results), but I’ve compared Niese’s fastballs per start, where the effects should be less. See my earlier article on niese this year, he’s clearly changed his two seamer at the least (It’s possible the four seam improvement is an illusion due to park effects) as it was no where as distinct last year.
LASTLY: Are you the John who hangs out around tom tango’s website as well?
Yeah the more and more I look at last year I’m convinced he did not throw the cutter last year. The horizontal movement is just too close………they all seem like one pitch (4 seamer).
I’m looking forward to seeing what Harry can do to adjust the numbers by park. I mean we could look at Citifield starts only if we wanted to compare one start to another but if Citifield is off for some reason, we still wouldnt be able to compare his movement with everyone else. We really couldnt even compared Niese 2008 to 2009 since Shea and Citi may have differed. I’ll definitely check out your earlier article. I really havent gotten the chance to look at Niese this season yet in terms of pitch fx anyways.
Yep. I love the book blog. I’m the same john lol.
Yeah looking back then the horizontal movement of the 4 seamer and what I had labeled cutter was awfully close so they probably were just 4 seamers.
At the time I was using K-means clustering algorithm to classify the pitches.
Now using spin and spin axis its alot easier to distinguish between pitches.
So over this good run for Jon Niese you say he’s using the cutter alot more? Thats encourgaging to hear. Maybe next time he gets called up, he’ll pitch well enough to stay.
i just want to say again that this post was awesome. cannot stress that enough.
I really wish they had pitch fx installed in the minor league stadiums as well. It would help so much analyzing the various pitchers thoughout the system.
agreed, even if it was just at AAA level
Not surprisingly
Looking at Fangraphs
Jon Niese run values on the fastball
2008 -3.3 runs or (-1.97 per 100 fastballs)
2009 -0.1 runs or (-0.08 per 100 fastballs)
its basically turning into a league average fastball. His curve has actually hurt him more this season.
Interesting. I think July and August will give us the opportunity to have a much more robust dataset on Niese.
I’d caution against using Run Values for this sample size…i don’t have the numbers in front of me (i’m on my grandpa’s computer, not my own) but iirc one of those curveballs resulted in a home run. Which has a huge effect on run value (+1.4, on average)…and Niese only threw a few curves this year (for some strange reason).
As such, run values in this small a sample (especially on the curveball and two seamer) are probably not a good sign of effectiveness for the future, just for that one start.
Oops my bad, it was a double. Fangraphs also appears to be combining the two fastballs for their run values (Baseball Info Solutions, not PitchFX is where they get their data for the non pitchfx stuff).
Still 22 pitches is a really really small curveball sample size.
Yeah I totally agree.
Also with run values, it doesnt take into account that the effectiveness of one pitch may be partially due to another pitch thrown. For instance, you might have a bad slider, but you may need to throw it anyways at times to set up other pitches.
Anything with Niese tho, until he makes more starts at the MLB level should be taken with a grain of salt.
This is a really interesting post. Optimism inspiring.
Given the 2008 data, you can see where some people, namely me, would be on about a straight fastball. Given the sample sizes, if this gets the benefit of the doubt as to not being an abberation, the question has to be “what’s he doing differently this year,” in terms of the 4 seamer itself more than his pitch mix, or “what’s he sometimes doing wrong that makes his fastball sometimes go from average to dead fish?”
The data makes it look like he has a chance to be effective mixing his 3 (counting the cut) fastballs, which bodes well.
Another question is, what’s the average differential between 2 and 4 seamers going to be? If any… looking at the two starts you can see a few pitches logged as 2 seamers clocked at 90, which seem high for a guy barely averaging that on his 4 seamer.
It’s interesting in that the average speed of both types of fastballs is the exact same (note fangraphs uses the gameday algorithm and thus its data on the 2 seamer is mixing the pitch with some change ups). This would seemingly suggest to me that Niese could throw the four seamer faster if he tried.
Of course he throws a ton more four seamers (something like a 5-1 ratio of 4 seam to 2 seamers this year), which means he might be throwing much harder on the two seamers rather than the four seamers, where he maybe conserving energy.
It’s something to look at in future data.
You’d probably expect a 1 or 2 mph difference but I do think that some pitchers throw it at around the same speed. I think when I did Pelfrey last year it was only about 0,5 mph difference.
He could be throwing it harder tho. It would be interesting to see if he gets called up and makes more starts if there would be a difference or not.