Quick f/x: Drabek's curve

The Yankees will face Kyle Drabek this afternoon.  It will already be the third time they have faced Drabek in his young career, though this will be their first meeting at Yankee Stadium.  Drabek will pick up plenty of groundballs (55% since his major league debut) and whiffs (25% of swings), but he's struggled to get the ball over the plate (12% walk rate, 43% ball rate, 33% zone rate).  I'm going to make this pretty brief (you can read some more about Drabek here), but I would like to check in with Drabek's curveball, which looked to be his primary out-pitch when he came into the league.  He's not using it as much as he was last year.  Game by game pitch selections are below; there's a slight overall decrease since last year.  


click to enlarge

Drabek has also changed how he uses his curve.  Ponder these metrics from Drabek's 2010 starts compared to 2011:

	#	Zone Rate	Swing Rate	Chase Rate	Watch Rate
2010 Curveball	38	.237	.526	.448	.222
2011 Curveball	44	.364	.432	.393	.500

The sample size is small, but a trend is pretty clear - Drabek is throwing more curveballs over the plate than last year and is able to pick up more called strikes.  Also of note is that his curve averaged 84 mph last year and is only at 81 this year.  It looks like it's also getting some more vertical "drop," but I won't say too much on that before working out the calibration issues at Rogers Centre (they've been pretty severe, sometimes off by more than five inches).

Drabek's curve is by far his best pitch in terms of run values (both expected and actual), and since last year has garnered a swinging strike on 59% of swings (that kind of number would be up at the top of the league over a full season, but we only have a sample of 39 swings).  He seems to be pitching to contact more this year by throwing a much higher percentage of cut fastballs, but it's always good to have a power curve at your disposal when you want to use it.

Soriano's velocity is up

I think it would be fair to say that Rafael Soriano hasn't exactly made the best first impression in the Bronx.  Although his performance has been sub-par, one early concern about Soriano seems to have been alleviated:

The graph, which starts on March 31, 2011, would imply that for the most part, Soriano's velocity has been on the rise.  In yesterday's outing against the White Sox, Soriano was throwing harder than he had all season, hitting 95 on the gun with both his four-seam fastball and his sinker.  


Note that the cutters on the graph might just be four-seamers with some extra movement on them; I'll need to look at more data to make a more precise split.

Notes on Humber

The White Sox make their first visit to Yankee Stadium tonight, and on the mound will be Philip Humber.  Humber was part of the Mets' package in exchange for Johan Santana, and I think it would be fair to say that his career hasn't panned out exactly as many would have hoped.  He's had several stints in the big leagues dating back to 2006, and he hasn't had a whole lot of success.  After being taken third overall in the 2004 amateur draft, he posted good numbers in the Mets' minor league system; following the trade in January of 2008, Humber saw his strikeouts decrease and his walks increase, and after the 2009 season he was let go by the Twins.  He was picked up by the Royals and had a pretty good season for Triple A Omaha last year with much improved control (4.9% BBHBP rate, the best mark of his professional career).  Despite his improvements, the Royals released him and he was signed by the White Sox in January.  He's now in the rotation after starting the year as the team's long reliever, and he's expected to stay where he is until Jake Peavy returns from his arm injury.  

• Humber has worked mainly with four pitches this year.  He throws a four-seam fastball that's typically around 90-91, a curveball around 80, a changeup in the mid-80s, and a mid-80s cutter that he learned this winter.  The cutter looked more like a traditional slider in his most recent start.  Humber may also throw a few two-seam sinkers, but not many.
• He's about average in terms of getting ground balls (44% for his career, 43% this year).
• He gives up plenty of contact (16% whiff rate so far this year).  The four pitches I've ID'd are all below average at getting swings-and-misses.
• His control has been pretty good so far (fewer than 3 base-on-balls per 9 innings).  His career mark is 3.6.

Humber's track record isn't terribly impressive, but this chance with the White Sox might be what he needs to rejuvenate his career. 

Sabathia's velocity against the Orioles

The YES radar gun has been improved since 2009 (no more 95 mph curveballs), but it was definitely giving some funky readings during tonight's game against the Orioles.  In the 8th inning, it had CC Sabathia throwing at 98-99 mph, which is well out of his normal range.  PITCHf/x data has him throwing harder than usual, but nothing like the YES gun:

	#	Min	Average	Max
Fastball	48	90.8	93.6	95.9
Sinker	18	90.3	92.6	96.0
Slider	23	79.6	81.8	83.0
Changeup	13	85.3	86.3	87.0
Curveball	1	77.4	77.4	77.4

Note that Baltimore is running faster than most parks and, as of last week, was on average about 0.8 mph faster than Yankee Stadium.


UPDATED: For some more context, compare today's numbers from those over his first four starts:

	#	Min	Average	Max
Fastball	209	90.3	92.7	94.9
Sinker	60	90.1	91.9	93.6
Slider	73	78.5	81.1	83.2
Changeup	71	84.0	85.8	87.5
Curveball	27	76.4	78.4	79.7

Sabathia's curveball classification can be kind of weird, as his curve has virtually the same movement as his slider and doesn't have a whole lot of velocity difference.  But looking at the numbers for the fastballs and changeup, I would say that Sabathia's velocity wasn't a whole lot different today from the rest of the season.

Colon and his two-seamer head to the rotation

With Phil Hughes injured/pitching poorly, the Yankees will give Bartolo Colon his first start in a Yankee uniform (and his first start overall since early 2009) this evening.  You could argue that Colon has actually been the team's best pitcher so far, as he's posted a team-leading 2.70 xFIP and 10.3 K/9 over 11 1/3 innings (4th most on the team) in long relief of Hughes.  He's shown four pitches so far, mainly relying on a combination of two and four-seam fastballs:

	mph	pfx_x	pfx_z
Two-seam fastball	89.9	-9.4	+5.0
Four-seam fastball	92.0	-4.8	+10.3
Slider	81.9	+2.0	+1.4
Changeup	82.1	-8.3	+6.7

The two-seam (45.7%) and four-seam (35.9%) have been his main weapons, though he's thrown a few sliders (12.5%) and changeups (6%).  Despite the 13 strikeouts, Colon's only garnered 13 swinging strikes - over 73 swings, that's a whiff rate of .178, which is below average.  Despite the small sample size, here are the whiff rates for Colon's pitches so far:

	Swings	Whiff Rate
Two-seam fastball	27	.074
Four-seam fastball	25	.320
Slider	12	.167
Changeup	9	.111
	73	.178

The four-seam fastball has looked very good in the whiff department, but in two-strike counts, it's been the two-seamer that's been most reliable for Colon.  A lot of his sinkers have been on the glove-side (inner to lefties, outer to righties) half of the plate, and he's been able to fool hitters with it in the strike zone with a watch rate of 50% (league average is around 37% for sinkers).  Of his 13 strikeouts, six of them have come on the spill-over, the "Greg Maddux pitch."  The graph below shows those six pitches (including some possibly generous calls), overlaid on an approximated strikezone.  Also, you can watch these videos to see some of Colon's strikeouts this season.

Colon has shown impressive command over his first three appearances, and if he pitches well, I wouldn't be surprised to see him hang around the rotation for a while given Hughes's injury and Ivan Nova's ineffectiveness.  I think the first big question is whether or not he will be able to hold up over a full start - questions about his durability were what landed him in the bullpen in the first place.  Also, will the spill-over fastball lose effectiveness as hitters see it more?  Will the free-swinging Blue Jays refuse to take it for a strike tonight?  I guess we'll find out.

Rivera's wild one

We all know that even Mariano Rivera has bad days sometimes.  Still, it was remarkable to see him throw a wild pitch on a 3-0 count.  It was just Mariano's third wild pitch since 2008 (A.J. Burnett already has six in four starts this year).  It crossed the plate at 0.26 feet above the ground, which is the lowest cutter Mariano's thrown according to the PITCHf/x data I have (regular season games going back to 2008).  The only pitch he's thrown lower was a sinker that he bounced for a wild pitch in an August 15, 2008 game at home against the Royals.  That pitch's vertical plate location was -0.14, which means that it bounced before reaching the plate.  Coincidentally, that wild pitch also scored a run from third base.

2010 League Averages by Pitch Type

I have been wanting to do this for a while, and I've finally gotten around to it.  The tables below show the league averages by pitch type for many of the metrics I typically use here. 

	#	Swing Rate	Whiff Rate	Zone Rate	Chase Rate	Watch Rate	Foul Rate	Ball Rate	nsCall Rate	RV/100	xRV/100
Fastball	262,588	.444	.162	.479	.268	.364	.431	.356	.360	-0.14	-0.21
Sinker	130,440	.441	.122	.459	.284	.374	.381	.362	.352	+0.00	-0.02
Slider	98,164	.474	.313	.419	.345	.347	.314	.371	.295	-0.73	-0.76
Changeup	79,370	.499	.288	.372	.363	.273	.297	.386	.231	-0.28	-0.15
Curveball	63,580	.396	.281	.403	.294	.453	.317	.413	.316	-0.26	-0.34
Cutter	39,796	.488	.209	.456	.318	.309	.387	.352	.312	-0.36	-0.42
Splitter	9,261	.515	.352	.352	.392	.257	.304	.390	.197	-0.82	-1.04
	697,108	.453	.208	.445	.302	.358	.373	.368	.327	-0.24	-0.28

	GB Rate	FB Rate	LD Rate	PU Rate	HR/FB	wOBAcon
Fastball	.385	.329	.196	.091	.103	.363
Sinker	.522	.245	.188	.045	.109	.353
Slider	.444	.296	.178	.082	.112	.345
Changeup	.482	.267	.180	.071	.116	.336
Curveball	.499	.256	.185	.060	.114	.346
Cutter	.448	.272	.197	.084	.106	.353
Splitter	.566	.230	.159	.044	.137	.338
	.450	.288	.189	.074	.109	.353

Yes, the run values are pretty funky.  They're out of date and the expected values don't perfectly add up to the actual ones.  This is something for me to fix for this year.

nsCall is called rate on non-swings, or called strikes over called strikes plus balls.  It's a different way of showing Harry Pavlidis's B:CS metric.  

Four-seam fastballs have the lowest HR/FB of any pitch.  Hanging breaking pitches get hit out the most.

Speaking of HR/FB, my values add up to be pretty different from BIS's at Fangraphs.

The "curveball" group contains a few eephus pitches as well (I have some from Vicente Padilla, Chad Billingsley, and Roy Oswalt).  

I also have knuckleballs (including Tim Wakefield, R.A. Dickey, and Charlie Haeger) and screwballs (Danny Herrera and Dallas Braden) classified, but didn't include them here because of the sample size.  Also, generic fastballs ("FA") that were neither specified as a four-seamer or sinker were not included.

UPDATED 7/27/11

Tillman on the hill

The Orioles come to Yankee Stadium for a three game set that, weather permitting, will begin tonight.  A.J. Burnett will go up against Orioles' youngster Chris Tillman.  Tillman's been around since 2009, but he's still only 22 (he'll turn 23 on Friday).  He was a highly-touted in the minors (Baseball America's 67th best prospect in 2008 and up to 22nd in 2009), and was a big piece in the 2008 Erik Bedard trade, but he hasn't really done a whole lot since 2009.  Here are a few things I've picked up from going through minor league stats, classifying his pitches through PITCHf/x data, and watching video:

• Tillman came up with a low-90s fastball, a mid-70s curve, and a changeup at around 80 mph.  He added a mid/high-80s cutter and started using it in the majors last July.
• He's tall (6-5) and he comes from a high arm-angle.  Not as crazy as Chris Young, but he still gets a pretty unique look.  His vertical release is similar to that of Mat Latos.
• As is the case with over-the-top motions, Tillman's fastball is typically pretty straight and fly-ball prone (.295 GB rate since 2009).
• His curveball gets a ton of "drop" (pfx_z typically equals -9 to -9.5 inches for him) and is similar movement-wise to the curves of Cliff Lee and Ted Lilly (though Lee's has some more horizontal "sweep").
• Tillman hasn't missed many bats in the majors (.161 whiff rate since 2009).  His strikeout numbers were strong in the minors (9.9 K/9 from 2006 to 2009), but dropped in 121 2/3 innings at Triple A last year (7.0 K/9).  He has improved his control, though, as his 2.2 BB/9 last year was the lowest in his professional career.
• His velocity is also on the decline.  He averaged 92 mph in 2009, lost around a mile and a half per hour in 2010, and so far this year he's only averaging 88.5 mph.  It's still early and pitchers don't usually throw their hardest this early in the season, but it's still alarming.

He's had a solid opening to his season (including a 6 inning no-hitter in his first start).  It's been rocky for him so far (5.84 FIP for his career, third worst in the majors since 2009), but he's still very young. 

Sox f/x III: Josh Beckett

With the series tied up at one game apiece, the Yankees will turn to their current ace, CC Sabathia, and the Red Sox will turn to their former ace, Josh Beckett, tomorrow at 8 on ESPN (more national TV!).  Beckett still has great stuff, but he's been hit hard recently (solid peripherals, but a 5.78 ERA in 2010).  He throws the same five pitches as Clay Buchholz.  Take a look:

	mph	pfx_x	pfx_z
Fastball	93.2	-6.9	+9.0
Sinker	93.1	-9.9	+5.2
Curveball	75.7	+7.0	-7.5
Cutter	90.3	+0.7	+5.6
Changeup	87.6	-8.8	+2.5

Beckett gets less drop and velocity on the curve than Buchholz, and some extra tail and sink on the sinker.  The biggest difference is the changeup - Buchholz's is slow and straight, while Beckett's just looks like a sinker with some velocity taken off.  Generic pitch results:

	#	%	RHB%	LHB%	Swing	Whiff	Zone	Chase	Watch	Ball	RV/100	xRV/100
Fastball	801	.379	.331	.411	.439	.139	.472	.255	.354	.356	+1.18	-0.82
Sinker	340	.161	.187	.143	.453	.123	.435	.313	.365	.388	+1.10	+0.46
Curveball	385	.182	.226	.153	.348	.291	.384	.257	.507	.436	+0.32	+0.09
Cutter	347	.164	.207	.136	.602	.273	.435	.443	.216	.308	-0.12	-1.31
Changeup	242	.114	.050	.158	.459	.306	.339	.313	.256	.434	+0.14	+0.74
	2142				.455	.205	.433	.304	.347	.377	+0.73	-0.28

	GB Rate	FB Rate	LD Rate	PU Rate	HR/FB	wOBAcon
Fastball	.444	.375	.153	.028	.148	.479
Sinker	.513	.167	.244	.077	.154	.392
Curveball	.491	.200	.255	.055	.182	.400
Cutter	.426	.265	.191	.118	.167	.439
Changeup	.523	.250	.182	.045	.182	.315
	.467	.278	.197	.058	.164	.425

Beckett struck out over 8 men per 9 innings last year despite lacking a legitimate swing-and-miss pitch (though the cutter is impressive adjusted for pitch type - the average cutter has a whiff rate of around 21%).  The chart below shows Beckett's pitch selection last year in each count.  Hitters see plenty of curveballs in strikeout situations:

Overall: he got burned once the ball was in play last year (high wOBAcon, HR/FB, and RV-xRV), but he has some quality stuff.  His cutter, in particular, looks good.  He can probably still be quite effective this year if he stays healthy and has some more luck.