PITCHF/x Scouting Report: Yu Darvish

is the newest high-profile Japanese player to ship out to the MLB. There will be plenty of narratives when the Yankees play Texas tonight. v. Darvish might be proclaimed the battle for the title of best Japanese starter, but what I’ll be watching is the pitching. The 25 year old right hander is undoubtedly the NPB’s most successful pitcher to head to the Major Leagues, and the Rangers invested over $110 million to have him play for Arlington. If you read the hype during his posting, Darvish has a huge pitching repertoire, and numerous sources said he had more than a few plus pitches. While I dreaded tackling this article due to his pitch arsenal, I also loved the challenge. As I’ve stated before, I’m still in the process of developing my own database for this series, and I’m heavily reliant on TexasLeaguers and BrooksBaseball. Turns out that TexasLeaguer’s GameDay’s algorithm for identifying pitch types absolutely rules, and although it’s not perfect, it did an incredible job with Darvish. Brace yourselves, the graphs are coming.

Pitch Repertoire

There was speculation that Darvish would showcase a pitch repertoire of possibly 8 pitches coming from Japan, but now that we’ve seen him pitch through 3 official games, he has shown us “only” 6 pitches. Thus far the 93 mph four-seam fastball has been his most used pitch, and is above average with its rising action. The 84 mph slider comes second, which has an absolute nasty break away from right handed hitters. The 91 mph cutter is his third most used pitch, and frankly I’m not that impressed with the numbers I’ve seen thus far, but we’ll talk about those later. The slow 75 mph curveball is another out pitch for the right hander, with huge movement into opposite side hitters. He also has a two-seam fastball that TexasLeaguers has a problem identifying and a split finger which acts as his changeup. For bonus point, the other two pitches I’ve seen discussed were a screwball he stopped throwing in 2006, and there was speculation about a true changeup.

In the image above we have Darvish’s release point from a catcher’s perspective. The ball is typically released from about 5.5 feet high, and although this would be considered low for someone with a 3/4 arm slot, it’s common with many Japanese pitchers who use the Tom Seaver drop-and-drive motion©. The pitcher does have a wide area of release that spans from around a foot to the left of the mound to almost 3 feet. Most of his fastballs are released from the inside of his release zone, with the slider released on the outside. I fear that hitters may be able to pickup on his slider based on  his release spot; at times he can throw it well over a foot from the four-seam fastball. He also releases the curveball from the highest point, which is typical for pitchers when attempting to achieve somewhat of an overhand slot. You can see the actual 3/4 arm slot and drop-and-drive motion in the picture to the right.

Here we have a visualization of the different breaks for each of his pitches from the release point on the right to homeplate on the left. In the top picture we see how far Darvish is able to throw from the mound. This not only helps him gain power from his legs, but the long stride allows him to release the ball closer to the plate and thus with slightly more velocity. Again, we see how he releases different pitches from different release spots, with the two-seamer closest into the mound, and the slider and splitter the furthest away. Half of his pitches have movement into right handed hitters, the four-seam, the splitter, and as usual the two-seam with the most. The other three pitches move away from right handed hitters, with the cutter having the least break, followed by the curveball and slider which have pretty amazing movement away from same side hitters.

In the bottom picture, we have a visualization of the pitch trajectory from a 1st or 3rd base side. Here, the four-seamer has rising action, the cutter has some slight last minute sinking movement, and the two-seamer has similar sinking action. The splitter and sinker also have comparable downward break, but remember that each does so with opposite horizontal movement. As always, the curveball has the biggest drop.

Here we have the horizontal and vertical movement of each pitch based on a catcher’s perspective. The origin of this graph is a no-spin pitch that conforms to gravity. We also have our first opportunity to correct the pitch types, so please note that the orange squares you see next to the triangle two-seamers have personally been corrected by yours truly. (Here is the original graph) Aside from being clustered together, the next graph will better explain why I made this correction.

Moving on, the two-seamer obviously has the most movement into right handed hitters, with anywhere from 5 inches to 10 inches of movement. The average vertical sink of this pitch is charted at 6.21 inches, which is slightly more drop than the 6.4 from last year’s league average. The rise on the four-seamer sits around 10 inches, which is significantly more than average. Finishing up the left side of the graph, you see a few splitters underneath both fastballs, which shows how the horizontal movement matches. With such similar movement and slightly decreased velocity, you can see why it works as his changeup and as a strikeout pitch.

The cutter is the first pitch to move away from right handed hitters. With it’s 6.24 vertical movement and 0.90 horizontal, most of these cutters have very little dramatic break compared to the four-seamer. Of course, I suspect that TexasLeaguers is again misclassifying many of the low four-seamers as cutters, but I’m less confident about that than the correction above. With the current classification, I am less impressed with the majority of the cutters, but you can see that some have broken quite wide at times, meaning potential is certainly there. Let’s go on to one of the most dynamic pitches I’ve seen doing this series, Darvish’s slider, which has an average horizontal break of 7.20 inches away from same side hitters and a -0.23 vertical drop. Compare this one , who had one of the best sliders in the game last year, who has a -0.94 vertical drop, but also only 2.38 horizontal movement. While movement doesn’t translate into success, from a scouting standpoint, its astounding to see such a pitch. His curveball also has some very strong break down and away from hitters, averaging a -7.90 vertical drop, but also a strong 7.35 inch horizontal break away from righties. Again, this horizontal movement is rare.

Here we have graphed the spin angle versus velocity of Darvish’s different pitch types. If moving from right to left, we can see how the splitter is thrown at the highest angle, but lets look at the two-seamer. Your average two-seam fastball is thrown at around 220-245 degrees from a right handed pitcher, indeed Darvish’s two-seam begins in that range, but TexasLeaguers begins to classify them as four-seamers before we go below 220 degrees. With the horizontal and vertical breaks matching up to these pitches in the graph above, and here the spin angle matching up, I am confident that he’s throwing a two-seamer here. Your average four-seam fastball is thrown at around 200-220 degrees from right handed pitchers, and you can see that at 220 degrees, there is a break between his two and four-seamer. Towards the lower end of the four-seamers you can see the cutter begins, and although the cutter is more of a subjective pitch, I usually classify them around 160-190 degrees. One could argue that the classified cutters thrown from about 180-190 degrees are four-seamers, but I think you’d have to check the video of every pitch in this case or ask Darvish himself. I’ll trust TexasLeaguer’s GameDay’s algorithm on this one as I’ll never be confident enough. Less complicated are the spin angles for the slider and curveball, which fall well below the rest of the pitches.

Where He Throws It

In this graph we have the location of each of his pitch types to right handed hitters, but please keep in mind that the two and four-seamer are misclassified. While Darvish likes to approach righties with pitches up and in or down and away, he isn’t afraid to throw his slider and four-seam for strikes. Likewise, he is able to throw his curveball in the strikezone, which can be a nice tool in the beginning of a count. If there is any pitch thrown specifically to a location, it would be the cutter which is mostly located down and away to catch a righties chasing. Although the sample size is small, I would be wary of splitters and two-seams down and in to jam batters that aren’t lose on the curveball or slider. According to the algorithm, his pitch selection against right handed hitters is 46.8% four-seam, 27.0% slider, 13.5% cutter, 9.0% curveball, 1.8% splitter, and 1.8% two-seam. (Which is likely closer to 10%.)

Facing left handed hitters, Darvish’s plan is to attack them away. Most of his fastballs are thrown on the outer half of the plate in order to get a hitter chasing the pitch as it breaks away. The cutter and slider, which have movement into this hitter, can be thrown for strikes, but he also appears to be hitting the down and in corner with the pitch. The curveball here is thrown mostly down and away as well. While the curveball and slider have such big movement into left handed hitters, I’m surprised he doesn’t attempt to jam them down and away more often. His pitch selection is 45.0% four-seam, 17.6% cutter, 15.8% curveball, 15.3% slider, 5.4% two-seam (which is likely closer to 15%), and 0.9% splitter.

When He Throws It

Yu Darvish’s Pitch Selection Against Righties

Count	Four-seam (R)	Slider (R)	Cutter (R)	Curveball (R)	Two-seam (R)	Splitter (R)
3-0	50.0%	0%	50.0%	0%	0%	0%
3-1	66.7%	0%	33.3%	0%	0%	0%
2-0	25.0%	25.0%	50.0%	0%	0%	0%
2-1	50.0%	50.0%	0%	0%	0%	0%
1-0	41.7%	25.0%	25.0%	8.3%	0%	0%
0-0	42.9%	32.1%	10.7%	7.1%	7.1%	0%
1-1	70.0%	10.0%	20.0%	0%	0%	0%
0-1	50.0%	28.6%	7.1%	14.3%	0%	0%
3-2	60.0%	20.0%	0%	20.0%	0%	0%
2-2	60.0%	20.0%	0%	20.0%	0%	0%
1-2	27.3%	27.3%	9.1%	18.2%	0%	18.2%
0-2	37.5%	50.0%	0%	12.5%	0%	0%

Here we have Darvish’s pitch selection in different counts to right handed hitters. Starting them off he is most reliant on his fastball and slider, throwing the combination more than 75% of the time. As he falls behind in the count, he relies more on his fastball and cutter to get strike one. As he gains in the count, Darvish continues throwing fastballs, but also mixes in his slider and curveball around 40% of the time. On an 0-2 count, he uses his slider as the primary out pitch. With such dangerous breaking pitches, 23.3% whiff rate with the slider and 10% whiff rate with the curveball, he could probably be throwing those more often for strikeouts.

Yu Darvish’s Pitch Selection Against Lefties

Count	Four-seam (L)	Slider (L)	Cutter (L)	Curveball (L)	Two-seam (L)	Splitter (L)
3-0	71.4%	0%	28.6%	0%	0%	0%
3-1	62.5%	0%	25.0%	0%	12.5%	0%
2-0	61.5%	7.7%	23.1%	0%	7.7%	0%
2-1	75.0%	12.5%	12.5%	0%	0%	0%
1-0	64.7%	8.8%	14.7%	5.9%	2.9%	2.9%
0-0	35.0%	23.3%	21.7%	15.0%	5.0%	0%
1-1	23.8%	33.3%	9.5%	23.8%	9.5%	0%
0-1	47.4%	15.8%	21.1%	15.8%	0%	0%
3-2	57.1%	7.1%	14.3%	14.3%	7.1%	0%
2-2	30.8%	7.7%	15.4%	30.8%	7.7%	7.7%
1-2	10.0%	10.0%	0%	60.0%	20.0%	0%
0-2	0%	14.3%	28.6%	57.1%	0%	0%

Facing left handed hitters, Darvish begins the count mixing in his cutter, slider, and curveball along with the fastballs. As he falls behind in the count, he again relies more on the fastball and cutter. While gaining a favorable count, he mixes the slider and curveball in, with his curveball as the favorite out pitch here. Against lefties thus far, it has been is only real out pitch with a 14.3% whiff rate, followed by a relatively low 8.8% whiff rate on his slider. Starting hitters out with breaking pitches is very rare to see, so it could be an issue of inexperience here. The right handed pitcher might be more successful against lefties if he doesn’t show them the curveball or slider until strike one or two.

Results
In his short career in the majors, he has a batted ball profile of 42.1% groundballs, 29.8% flyballs, and 28.1% linedrives. I doubt this will continue, Darvish’s wide repertoire of sinking pitches should ensure that he maintains a high groundball rate. Thus far he’s been highly successful facing righties, holding them to a slim .174/.321/.217 triple slash, but has been hit hard by lefties with a .313/.431/.396 triple slash. It’s hard to tell how his home and road split will look, but in his one start at Ranger’s Ballpark, he gave up 4 runs in the first inning, followed by 4.2 innings of 1 run ball. Facing the Twins and Tigers in his only other two starts, he went a combined 12.0 innings giving up only 2 earned runs. Although he’s posted a 7.1 K/9 and 6.6 BB/9, I don’t believe that these are indicative of the pitcher that he is. Take a look at the graph below of his strikezone thus far, and tell me that Darvish hasn’t been squeezed.

Prediction
Darvish has some incredible break on his pitches, but the question remains whether this translates into major league outs. He’s been very impressive in his last two outing, but the Yankees lineup appears to embody the worst possible lineup he could face. While his platoon split against lefties could be small sample size, his approach against them appears unexperienced. There is a lot of potential in his arsenal, but I don’t think he’s truly learned the major league game yet. With possibly 6 patient and hard hitting lefties in the lineup tonight, this will probably be the hardest start of his career. That said, there are very few pitchers with his type of stuff, and I’m jealous that he isn’t pitching in pinstripes right now. I suspect the Yankees will knock him out tomorrow, but with more experience in the majors, he won’t be an easy opponent for very long.