premium blogad:
blogads:
advertisement:
sitemeter:
|
|
|
June 11, 2009
If You Can Hit It, You Can Rename It
Recently, the team over at FanGraphs.com added Pitch Type Linear Weights to their already robust website. According to David Appleman,
Using this information, let’s take a look at which Mets have the most effective pitches. Please note that using runs above average per 100 pitches, “/C,” tends to favor relievers due to smaller sample sizes and the general nature of relievers, think UZR/150. While using cumulative runs above average favors starters as they have a great chance of cumulating more value, think VORP. Top 5 Mets Fastballs wFB/C Feliciano 2.61 Parnell 1.56 Rodriguez 1.27 Stokes 1.19 Maine 0.41 Bottom 5 Mets Fastballs wFB/C Green -2.26 Putz -1.63 Takahashi -1.61 Perez -1.57 Hernandez -0.71 That’s right Pedro “87.4 MPH” Felicano has the Mets best fastball per 100 fastballs. That’s not to say he should throw it more since it’s clearly blowing hitters away. Rather, the effectiveness of his fastball is due to the fact that he only throws it 53.3% of time while coupling it with a great slider and a changeup that he has been throwing twice as much than in years prior. As for the bottom five, well Takahashi and Livan seem to understand their fastballs are “show me” fastballs focusing more on their superior off-speed offerings, as for the others, well at least they got their health…erm…well, paying contracts. Top 5 Mets Sliders wSL/C Green 1.98 Putz 1.87 Hernandez 1.67 Pelfrey 0.71 Maine 0.38 Bottom 5 Mets Sliders wSL/C Stokes -3.70 Takahashi -3.07 Perez -2.41 Parnell -2.13 Redding -1.21 The top three sliders belong to pitchers that all made an appearance on the bottom five fastballs, so they have that going for them. I’m surprised to see Pelfrey and Maine on the list over Felicano. Last year, Pelfrey’s wSL/C sat at -1.92, so it seems he might have made some real progress with the pitch which is great to see. Stokes and Parnell are two top five fastballs that also make an appearance on the bottom five slider chart. The latter agrees with the scouting reports and all of our eyes as the two can throw gas but don’t really have a matching strong breaking pitch. Stokes is what he is at this point, but there is still hope that Parnell can make like Pelfrey. Takahashi and Ollie are back on the bottom five but one is about to redeem himself (can you guess who?) Note: Felicano’s slider is sometimes labeled as a cutter according to FanGraphs. His “cutter” stands at wCT/C of 2.77, meaning his slider probably belongs on the top of the list but I can’t really break the data down properly. Top 5 Mets Changeups wCH/C Green 10.26 Putz 9.18 Takahashi 6.83 Rodriguez 4.07 Santana 2.39 Bottom 5 Mets Changeups wCH/C Felician -8.04 Redding -3.98 Perez -3.13 Parnell -2.87 Pelfrey -0.30 Green and Putz’s numbers are probably a bit of noise, as their past performances do not indicate anything close to this level of ability. As we can see, Takahashi changeup redeems him from is lackluster showing thus far; as for Ollie, not so much. I’m pretty surprised to see Felicano so low since his changeup has really helped his game. What we can glean from the information is Felicano probably wasn’t throwing his changeup as much until this year because he realized it was a bad pitch and was getting hit. However, the use of the changeup appear to have made his other pitches more effective and is without a doubt worth the trade off, which I’m also sure he realized. Top 5 Mets Curveballs wCB/C Green 5.35 Redding 4.15 Hernandez 3.97 Stokes 1.26 Rodriguez 1.06 Bottom 5 Mets Curveballs wCB/C Perez -4.92 Niese -3.97 Putz -3.93 Maine -2.71 Pelfrey -0.44 Tim Redding makes an appearance on this top five cementing that every Met pitcher at least does something right, sans Oliver Perez, who somehow makes it onto the curveball bottom five. It’s a shock that Niese’s curveball is the second worst—granted he was unlucky in the BABIP department while he was up—but the curve is supposed to be his bread and butter and it looked good to me. I’m going to chalk the latter up to a small sample size issue. Also, it’s interesting that the owner of the worst fastball was able to top the remained of the pitch lists. Combing the aforementioned fact with his tRA, there is even more hope for Sean Green. There is a ton of information and guesses that can come from this data. Let me know what else you can deduce or disagree with in the comment section. 6 Responses to “If You Can Hit It, You Can Rename It”Leave a ReplyYou must be logged in to post a comment. |
|
Great work, Joe.
I just wanted to touch on one thing, something that you yourself touch on briefly when talking about Feliciano’s change.
Pitching is a balance, so this data is loaded with interaction effects. Just because a pitch is doing poorly, doesn’t mean it isn’t doing its job. Pitchers with great fastballs often have poor linear weight values attached, not because its a bad pitch, but because hitters are forced to look for them. The problem isn’t the fastball, but the pitches that go along with it (classic example: pitcher can’t throw the slider for strikes, hitters avoid the slider, look heater, and hit heater).
Similarly, you’ll often see the pitcher’s worse pitch (by the scouting reports) do quite well. The reason for that usually is that the pitcher almost never throws that pitch, and when he does throw it, it takes the hitter by surprise.
So this isn’t really just a demonstration of the pitcher’s ability with the pitch. It measures,
(1) The pitch’s actual “absolute” quality.
(2) Where the pitcher throws it.
(3) How the pitcher uses the pitch (as in how often, what situations, etc.).
So it’s fascinating data, but often a little misleading. What I’d love to do is log every pitcher’s data for this stuff, and maybe run some correlation studies (or maybe even some ANOVA’s).
Good points Alex, I agree wholeheartedly. Your right that I hinted at the “limitations” of the stat but did not fully disclose them. This statistic is a perfect example of a stat that is more than just a number. There really are so many assumptions, guesses and inferences that can be made using the data unlike a stat that is more static such as HBP.
I took the numbers more for face value for fun and there is some truth to the numbers, but the limitations are important, so thanks for adding important information to think about when using wPitch.
I agree with Alex. I mean if Pelfrey throws 80% fastballs, hitters tend to look for them and that might be the reason why his slider comes out good.
Another aspect thats kind of similar…..its about sequencing as well. You might have a pitch that you struggle with but you throw it in order to set up another pitch.
So we can look and see Stokes slider gets hit hard…..so he should ditch the slider….thats not necessarly true.
I think mixing up pitches and keeping hitters off balance is extremely important.
Yup, you’re right John. In regards to Pelfrey, he threw about the same percentage of fastballs last year. However, while the vertical movement on his slider has remained about the same, his horizontal movement has doubled from .9 to 1.8. Even visually, Pelfrey’s slider has looked improved and wSL/C supports that. It could be that Pelfrey’s slider isn’t better but that hitters aren’t used to seeing the increased movement and will adjust.
The numbers I presented are complementary data that can be used in conjunction to other information in order to fully explore a thought.
Another thing about him altho he’s been 80/20 this year fastballs vs offspeed lately he’s been more 70/30. I think that’s helped him strike out more batters
Last night tho I think he was back to mostly fbs
Not surprising, he did not k many last night
Last season I developed my own run values by count when i did my pitch fx articles On fangraphs forum I asked David to show there chart by count so I can compare