The Best Pitcher with a Bad Fastball
Yesterday Adam Wainwright won his league leading 16th game, strengthening his Cy Young case. Of course readers here know how dubious it is to look at wins as a measure of pitching talent. But, without a doubt, Wainwright has been one of the top pitchers this year, with a FIP of 3.33 and a tRA of 3.71.
Interestingly Wainwright’s fastball has been pretty poor this year, and he has succeed on the strength of his very good slider and curve. Those two pitches have saved over 32 runs, the next best breaking pitch combo belongs to teammate Chris Carpenter whose slider and curve have saved 21.8. As a result Wainwright throws his fastball only 50% of the time, which seems to be about the floor for how infrequently a pitcher can throw a fastball (if you consider a cutter a fastball and exclude knuckleball pitchers).
He throws his slider mostly to RHBs, from whom it moves away. Here are the locations of these pitches this year.
Perfectly clustered on the outside of the plate. It is not a huge-whiff inducing slider, only 27% misses per swing, compared to the top sliders which get in the in the over 40%. Instead it gets value from of out of zone swings (35%) and weak contact that results in lots of grounders (49%).
As I wrote about earlier his curveball was one of the best in the game. It still is, ranking second. It gets lots of out of zone swings (40%), while only getting 55% in zone swings. That means hitters are only slightly more likely to swing at it in the zone than out. Showing how deceiving it is and resulting in called strikes and swinging at balls. On top of that it gets lots of whiffs (33%) and grounders (59%). An incredible pitch.
Wainwright has below average fastball. It generates few whiffs and few out of zone swings (although it does get a good number of grounders). So he throws it just enough to get ahead in the count and throw his devastating break stuff.
Dave Allen's other baseball work can be found at Baseball Analysts.
Apparently LaRussa (or someone) is a lot better than I used to credit them for. Someone in the Cards organization is making average-ish arms become extremely effective. Not to mention how much more effective Smoltz looks (blah blah blah Dave Cameron called it blah blah blah, most knowledgable Sox fans don’t attribute Smoltz’ struggles to a lack of stuff, it was a lack of stamina that seemed to be the issue).
And of course that dastardly SSS
dave duncan is the magic man. supposedly carpenter saw smoltz tipping his pitches, which has led to he resurgence. apparently no one on the sox pitching staff pays attention.
I know, so much for John Farrell making a good manager some day.
At least Epstein’s hands are clean on it, not his job to dissect their performance. As a Sox fan, I’m wicked annoyed that no one picked up on it, since I’m convinced there’s no way he goes from getting his ass handed to him 6 out of 8 starts (one of which was a 52 pitch, 4 inning game) to 2 straight shutdown performances as soon as he changes laundry.
Sucks that I now can’t bring myself to criticize LaRussa & Co, they know how to fix problems.
Pitchers often perform better when they’re sent down to the minors.
I think it’s a bit unfair to say Wainwright is an “average arm”. He’s had plus stuff for a few years now. Not devastating, but certainly much better than “average”, and he’s had a wipeout curveball since before he whiffed Carlos Beltran with it to win the NLCS and Brandon Inge with it to win the WS in 2006, as a pretty effective closer.
To be honest, I think Dave Duncan has had little to do with Wainwright’s excellent season. He’s been slated as the de-facto ace for a couple of years, did OK in 07 in his first year as a big-league starter, and was only hurt last year by a freak finger injury that kept him out for more than 2 months. This sort of improvement was fairly predictable, I’d say.
Maybe average is the wrong word. But it’s obvious Wainwright is getting outs based on being crafty more than just having Verlander-esque “stuff”. If you get my drift.
Actually Wainwright struck Inge out with the Slider to end the ’06 series further evidence on how effective both his breaking pitches can be.
Joe
You’re saying he doesn’t have a big fastball…but I’ll put Wainwright’s breaking stuff up there with anyone’s. It’s not crafty. Glavine, Maddux, Moyer are crafty. Wainwright is nasty, it’s just that his fastball isn’t 96.
Being a noob to this site, but looking to enjoy it, it would be way helpful if many more acronyms appeared in the glossary. For instance, this article uses both FIP and tRA, which are not in your glossary. Oh, please help.
Jeff,
yeah new stats are added quicker than the glossary is update. I will see about getting it updated. Until then on the player pages at the top of each stat section there is a button for ‘quick glossary.’ There you get a quick description of each stat. It looks like tRA is not in there, but I think most everything else is.
FIP is a stat created by Tom Tango. Eric from this site actually explains it here: http://www.ehow.com/how_2288361_calculate-fip-statistic.html, but as you can see, it’s not too bad. (13*HR + 3*BB – 2*K)/IP + 3.20.
tRA is more complicated, but essentially it tries to factor in ALL balls. In other words, a pitcher’s peripherals may be good (good K/BB, low HR rate), but if he tends to give up a lot of LD’s, it would reflect in his tRA. A brief description is found here: http://www.fangraphs.com/blogs/index.php?author=17
The detailed explanation of tRA from its creator is available if you click on my name.
Interesting article, I have one question though: is the assumption that Wainwright uses his fastball more to get ahead in the count an assumption, or does he have a higher FB% on 0-0, 1-0, 0-1?
If your DB could do it quickly, could you query wainwright’s FB% by count (or pitch # to a batter).
This is not a pointless exercise, intuition agrees with the assertion that the FB is a “count pitch”, and the slider/curve are “out pitches”, and it would be interesting to see if the stats confirm this.
One more question: are the linear pitch type weights used here adjusted for the expected run value of a pitch in a specific count/situation?
Here are Wainwright’s FB% by count. I think he follows the general trend of more fastballs when behind in the count (and early in the count) and less when ahead. Just his numbers are below average, it would be interesting to see if they are below average across the board (my guess) or just in particular counts.
0-0: 62%
1-0: 69%
2-0: 74%
3-0: 91%
0-1: 45%
1-1: 45%
2-1: 53%
3-1: 89%
0-2: 25%
1-2: 24%
2-2: 40%
3-2: 53%
One more question: are the linear pitch type weights used here adjusted for the expected run value of a pitch in a specific count/situation?
I am not entirely sure I understand your question, but here is the link to how the pitch values are calculated. If you that does not answer it you can follow up any questions you have here.
Very cool, thank you mr. Allen, I appreciate the query. The numbers indeed support your assertion that he uses the FB to get ahead (and get back when he’s behind) as a “count pitch”, and the SL/CU are the out pitches. Very cool, with such nasty breaking stuff (and throwing it 75%+ in 0-2, 1-2 counts), opposing batters must have terrible wOBAs in those situations, much much worse than league average I would guess.
In regards to my other question, I wondered if the linear weights pitch value took into account the run expectancy for the count/bases-out state of the event (pitch). From the link you posted, it is very clear that that is exactly what the linear weights are.
2006 NCLS, bottom of the ninth, bases loaded, and Cardinal killer Beltran at the plate with the Mets trailing 2-1 with two outs. Wainwright gets two strikes on Beltran and decides he’s going to throw him the most hellacious curve ball ever. Beltran can’t get his bat off his shoulder as Tim Welke puches him out and sends the Cardinals to the WS.
Now that was an out pitch. And I’m still wondering what Beltran was looking for.
Who knows what Beltran was looking for, but that curve was savage. Beltran supposedly said afterwords that he wouldn’t have hit it anyway.
I think some other information is necessary here.
Wainwright’s FB was a 1/2 win pitch for him last season, according to pitch values. He was also throwing it 7/10 of a MPH slower and throwing it 2% more often. So does this mean it’s an ineffective fastball now? Well, yes. But that does not mean it always has been or will be. According to pitch f/x the thing still moves a lot, not that much differently than Carpenter’s (it just happens to travel slower).
To me, if the “stuff” (movement and speed) is still fine then the location is likely off. Perhaps that’s the issue. But calling the pitch plain “bad” is a bit unfair in my book.
Might be sample-size issues too. Doesn’t take more than a couple of flyballs to turn into HR to turn a slightly-above-average pitch into a slightly-below-average one, especially if it’s only been thrown 60-odd% of the time (and, crucially, more often in hitters’ counts).
Yeah definite sample size issues. Those numbers are not ‘luck’ adjusted so any hr/fb or babip issues could make the fastball look worse than it is. I would probably take the career average numbers as the best guess as how good his fastball is. So in Wainwright’s case -0.13 per 100 pitches is pretty close to average. I probably should have titled it ‘A Very Good Pitcher with an Average Fastball’ it just doesn’t sound as good.
Does the pitch data account for the count for when the pitch is thrown? The data you posted above showed that Wainwright does throw his fastball predominantly early or when he is behind in the count–eg. the best hitting counts. This would account for poor success throwing his fastball. You could probably make a good argument that he should mix up his 3-0 and 3-1 pitch selection a little more than just throwing 90% fastballs.
Maybe his fastball is “bad” because his breaking stuff is so dominant. Seems to me that a pitcher with a good slider and devastating curve will have batters sitting on the fastball a lot, simply because that’s the only pitch they can hit. Not that his fastball is bad, but every hit he gives up is on the one pitch (because it’s the only hittable one).
Has Wainwright been throwing his changeup this year? How effective has it been? He has one (It is the pitch he got Inge to swing on in the WS).
Wainwright struck out Inge with a Slider. He does however use an effective changeup this season when his breaking stuff lacks – Not very often
Damn, I wish I had seen this article earlier. His average fastball for the year says 91 mph, and he usually tops out at 94 still. As a closer he’d usually throw one 96 mph fastball and then live with his phenomenal curveball or slider.
Since when is 94 mph considered a bad fastball? It’s not like he can’t throw in the mid 90’s when he has to.
This season his fastball on average is faster than both Johan Santana and Cole Hamels. And unlike those two who are showing decline, he’s struck out 212 guys and will get a lot of Cy Young consideration.
94 can be a terrible fastball if it’s dead straight, or it can be an amazing fastball if it has a ton of action on it. Regardless, Dave used a results-oriented analysis of the fastball, and has readily admitted its poor effect could be the result of the vagaries of batted balls over a small sample space.
It is inconsistent with the anatomy and neural control of the human vocal tract. ,
Question on taking into account pitch values for one season, does it not suffer at the hands of small sample size-ism?
Let’s say Pitcher X throws 50% fastballs, 100 avg pitches/start, 34 starts a year, 10 seasons.
1700 fastballs in one season, out of a potential 17,000 fastballs. Could it just be a ‘slump’ for a particular pitch? In the above example, It’s only accounting for 10% of his total career fastballs – a blip on the radar.
Now if it were backed up with reduced velo, movement etc maybe it would lend more credence but it seems to have the same limitations for all stats with SSS?