Monday, April 28, 2014

Masahiro Tanaka: Start Five

Masahiro Tanaka's fifth start of the season was his shortest so far, but still relatively successful, as he went 6.1 innings and gave up two runs, including a homer, and struck out 11 and walked 4 against the Angels. Through five starts this year, Tanaka has given up 10 runs, but 5 homers. He has a 2.27 ERA and a 2.91 FIP.

Like usual, let's start by looking at the home run Tanaka gave up, this time to David Freese, on the first pitch of an at-bat.

Freese is a right-handed hitter, so the fastball was actually up and away from him. For comparison, here is how Tanaka pitched him in the other at-bats.

Tanaka went middle to high and away twice more, and gave up contact again, though for an out, and got a foul ball. When he stayed inside, he was more successful.

Here is Tanaka's spin and speed chart from the game:

Perhaps this was just affected by the park, but it seems that there were more high spin splitters in this outing. Let's see if there was any major differences in how he located in this outing compared to other outings:

The curve was still relatively high, but this time he really struggled to get the slider down. He also couldn't get the sinker glove side enough to get inside the strike zone, though his splitter was still located well. Now, here is an average result location graph for comparison.

Tanaka was more likely to give up a foul ball when he went furthest glove side. His furthest arm side pitches were the most likely to give up runs. His whiffs, yet again with the splitter, came low and arm side. It appears that they came further arm side than his average splitter as well, suggesting that his furthest arm side splitters were the best. Strangely, pitches down in the zone were less likely to turn into outs, while his outs and called strikes were basically the same, arm side and middle.

This graph will show whether or not Tanaka was able to successfully locate away from both lefties and righties:


This graph not only shows how low he threw on average (as the entire graph is an approximation of the strike zone), but it also shows that he never really got the ball glove side consistently. He was always arm side, which was good against lefties, but not as good against righties.

Tanaka's groundball percentage was the second best out of his five starts, and this time, there was a big difference between the two

Tanaka got groundballs when he stayed arm side and low, like the splitter, but gave up flyballs when he went higher and glove side, like the slider or fastball. Digging further into location, the next two graphs will look at how he located by the count in the at-bat. First, let's look at his average locations by balls in the count:

With no balls in the count, he was more likely to go higher in the zone, but the deeper into the count he got, he went lower. The lowest, and only glove side location on average, was with two balls. With one ball and three balls, he located nearly exactly the same, which seems a little strange. Here is how he located by strikes in the count.

With no strikes, he located just about everywhere, which brings his average to be right down the middle. With one and two strikes, he really attacked low, bringing his splitter more into play.

Lastly, let's see if he located any differently based on what inning it was:

In his short time in the 7th, he threw the lowest and the most glove side of all his innings, though it is notable his other glove side inning was the beginning of the game. The third was a strange inning, in that it separated his consistent 2nd, 4th, and 5th innings. The 6th inning was odd in that it was his highest zone and extremely arm side.



Sunday, April 27, 2014

Breaking Down Shohei Otani's Most Recent Start

With Yuki Matsui now in the Ni-Gun, I wanted to continue my breakdown of NPB starts. So until he gets back up, I will look at individual starts of notable NPB pitchers, with this one being about Shohei Otani's start on April 27th against the Chiba Lotte Marines. Through four starts this season, Otani has struck out 28 batters and walked 7 with a 2.83 ERA in 22 innings. His start against the Marines was his worst so far, as he still struck out 9 in 7 innings and walked just 1, he gave up 2 home runs and 5 runs, all earned.

Let's start looking at his start by looking at his velocity and pitch count chart:

It appears that Otani maxed out at the end of his outing with his fastball. At the beginning, he started slow for him, and built up to around the 40-50 pitch mark before decreasing again. Once he got to about 100 pitches, he started throwing as hard as he could, with both his fastball and his splitter/fork. His slider velocity basically moved the same way his fastball velocity.

So since this wasn't a great start for Otani, let's break down his results:

It seems that he gave up a lot of fall balls, but he had a good strike percentage and got plenty of called strikes. His whiff percentage was only about average. Looking at the pitches he threw in the strike zone, it appears that if anything, Otani threw too many pitches in the strike zone.
Here is where Otani liked to locate his pitches in the outing:
Otani threw a great number of pitches low and to the glove side. However, Otani's second biggest location was right down the middle, perhaps confirming that he threw too many hittable pitches (as the two home runs show). Middle arm and high glove were his least frequent locations. Otani only threw three different kind of pitches, and his pitch breakdown may give us an insight on why and how he located:

His fastball percentage was perhaps a little low, and he threw a lot of sliders, which may be why he located so many pitches low and to the glove side. One of the home runs he gave up was on the fork (splitter) and another on a fastball. Looking further into his pitch selection, the following two graphs breakdown how he got groundballs, and how he got flyball. As the second graph shows, he gave up more flyballs than groundballs:




Otani was more likely to get groundballs off the fastball, and flyballs off the slider. His forkball was also more likely to get flyballs.

Since, like most pitchers, Otani's main pitch is his fastball, let's compare the pitch to his overall results and locations.

Clearly Otani located the fastball to the low glove side less than his slider. Surprisingly he threw less pitches down the middle with the fastball than his other two pitches. Middle glove and low arm side were two locations that saw increases, as well as high arm.

Otani's fastball results:
The fastball is where a good number of his foul balls came from, and his strike percentage was higher with the fastball. His whiff percentage was lower with the fastball. The following graph also shows that he threw more pitches in the strike zone with the fastball than his other two pitches:

Thursday, April 24, 2014

Yuki Matsui: Start Four

Yuki Matsui's fourth start of the season did not go well, as he walked eight batters and struck out six. In 127 pitches, he went five innings and gave up five runs. He was immediately sent down to the ni-gun after the game.

We will start the post the same way all the Matsui posts have been started, by looking at his velocity:

Matsui's velocity clearly declined as the start went on, especially towards the end, when he clearly shouldn't have been pitching anymore. Even his change and slider velocity dropped towards the end.

Matsui's big problem was obviously throwing strikes, as this graph shows what percentage of pitches he threw in the strike zone

Throwing less than a third of his pitches for strikes, it was clear that something was wrong with Matsui, or at the very least, he didn't belong in the Ichi-gun for the time being.

Looking at Matsui's results, we see that he still got whiffs despite his control problems:

Matsui got much more flyballs than groundballs, and got the same amount of called strikes as foul balls.

What may be helpful when discussing his problems with control is to look at is where he generally located all of his pitches:

Again, his favorite location was low and glove side, though he threw a lot of pitches high and glove side as well, meaning he was probably just yanking the ball overall. He also threw high more instead of low, while the majority of pitchers throw low more than high.

To further evaluate his outing and control problems, I think we need to look at his individual pitches. First, let's look at his overall pitch selection:

Notice how it appears that he has dropped the curve, at least in this outing. He stayed with the change more than the slider, and had a relatively normal amount of fastballs.

Looking deeper at his fastball, let's start by looking at the percentage of them he threw in the strike zone:

This rate is higher than his rates on sliders and changeups, but is still not an acceptable rate.

Here is where he generally threw his fastball location-wise:

This is where the high percentage of high glove pitches come from, though there was a good portion of low glove and high middle pitches as well.

Here were his fastball results for the game:


Matsui's whiffs didn't come from his fastball, and a higher volume of called strikes and fouls came the fastball.

I wanted to look at his locations of his other pitches, so let's start with the slider:


 This is where his high percentage of low glove pitches come from, though some of them stayed up a little and were middle glove.
 
His changeup locations:

Not surprisingly, this pitch was mainly located on the low arm side, with a good balance in the other zones. So really, the slider and changeups are in traditional locations, they just aren't finding the strike zone.

Tuesday, April 22, 2014

Masahiro Tanaka: Start Four

Masahiro Tanaka made his fourth start of the season on Tuesday, pitching in Fenway Park and going seven and one thirds of an inning, walking no one and striking out seven. He gave up two runs, both solo homers, to go with seven hits. Tanaka has now given up one home run per start (two against the Red Sox, none against the Cubs), and let's start this post by looking at the two home run at-bats, which actually came right in a row.

The first was to left-hander David Ortiz:

Other than one low and in fastball out of the zone, Tanaka stayed away from Ortiz, but notice that the home run ball was closer to the middle of the plate than the other three pitches.

The second was to right-hander Mike Napoli:

This time, Tanaka continued to go arm-side, but against Napoli, that meant he was going inside. The pitch that was most up and in, what appears to be a sinker that didn't sink, turned into the homer by Napoli.

For comparison, here is how Tanaka pitched David Ortiz in the non-home run at-bats:

Tanaka mainly kept the ball up and away, but notice that he got two swinging strikes by keeping the ball low and in the middle of the plate, similar to where the homer had been hit.

And here is how he pitched Napoli in the other two times he faced him:

Tanaka was much more successful at keeping the ball away in both at-bats, but got burned again by going up and in.

Here is what Tanaka's overall pitch selection looked like from a spin and speed perspective, labelled with MLBAM tags:


One thing that stands out is fewer curveballs. Also, Tanaka threw more four seamers than sinkers. However, the most prominent thing was his velocity. It seems that he threw more pitches over 95 MPH than he had in any other start. This is probably not park related, as Fenway Park is actually known to be a "cold" park when it comes to velocity. For whatever reason, it appears Tanaka was just throwing harder.

Since we looked at location in individual at-bats, let's compare how he located each pitch on average:

Again, I think we see why he went to the curve so rarely, as he still hasn't got it down like you would want. However, he kept all his pitches down on average, and had good horizontal differential with his sinker and fastball. His splitter stayed below the strike zone on average and the slider was in a decent spot.

For comparison, here is his average result locations:

 Not surprisingly, he had more success going glove side, showing that the slider and fastball were effective. His sinker really seemed to hurt him against Boston.

Tanaka's groundball rate wasn't bad against the Red Sox, but he fell short of the 50% mark that he posted in the first three starts. Here is where he located the average pitch that was hit on the ground and the average pitch that was hit in the air.


This time, there was basically no difference on average. Next, continuing to look at pitch results, I wanted to do something different, giving a visual look using movement, something we haven't looked at since Tanaka's first start. So here is a simple average strike, ball, and contact movement chart:


There seems to be a considerable difference when it comes to movement and results for Tanaka, at least in this start. An average movement chart with each pitch type will most likely demonstrate which pitches were the most effective:


This again seems to support that his fastball and slider were effective and his sinker was not. Notice that his splitter and slider have low vertical numbers, and
his balls and contact pitches had higher vertical numbers on average.

Thursday, April 17, 2014

Masahiro Tanaka: Start Three

Masahiro Tanaka's third start of the season for the New York Yankees was his best so far, as he went a season high eight innings against the Chicago Cubs, and gave up no runs, striking out ten and walking just one. He kept the ball in the park and only gave up two hits, though his groundball rate was his worst out of the three starts. Here is his speed and spin chart from the start, labelled with MLBAM tags:


His splitter has two different spin modes, one cluster that spins more than any other pitch, and another than spins like the slider. This may not mean anything, but to make sure, let's look at the speed and spin chart with only the pitches labelled as splitters, and labelled with results.


Clearly Tanaka was having a lot of success with his splitter no matter the spin and speed, though the harder ones in velocity seems to have been the most successful. In previous starts, Tanaka left a few pitches up and got hurt, so let's look at his average pitch locations to see if he did a better job of keeping the pitches down:


Again Tanaka couldn't get his curve down, and he couldn't get it glove side enough to be a strike most of the time. We see the same classifications that we saw in his last start, and I mentioned I didn't like the sinker tag, and according to the MLBAM tags, this was more common than the four seam fastball. The four seam and sinker/two seam fastball are both traditionally located, with the latter being lower and arm side. The slider location is still sketchy. He couldn't get it as low as you would want it, nor as glove side on average. His splitter location was great, and that seems to be what carried him in this start. Next, let's take a look at his result locations:


When he got the ball low and glove side, that is when he was most successful, while the further arm side or higher he got, the less successful he was. This really speaks to the curveball and sinker not being effective, while the splitter being extremely successful. Using average release point data, let's see if he is releasing each pitch in the same spot, or if he is being consistent.


There aren't huge differences in his release points, but there are some subtle differences that don't necessarily support a simple narrative. His curve was released closer to his body, and higher, but the closest one to it was his best pitch, the splitter. One interesting way to look at how a pitcher progresses through the game is to look at the average release points by inning:


Tanaka was more consistent from inning to inning than from individual pitch, with his biggest differences being the first and seventh inning, which were very similar to each other.

Moving on to the subject of the fewer groundballs, let's look at the average locations of the pitches that were hit in the air versus the ones that were hit on the ground.


Strangely, the pitches he got groundballs on were more arm side, while his flyballs were more gloveside. Not surprisingly, the lower pitches were the groundballs. Keeping on the subject of location, let's do a couple of things we did last time and compare them to his last start, starting with how he located pitches to left-handers and right-handers.


This time, Tanaka didn't pitch lefties and righties much differently. Instead of keeping the ball away from both, he kept the ball inside to righties on average. Like last time, let's look at his location by pitch count, but this time I'll break it up into two graphs, the first being the ones with strikes in the count, followed by one with balls in the count.


Not surprisingly, he was more likely to go glove side with two strikes, though Tanaka pitching lower on average with one strike was surprising.

The more balls Tanaka threw, the more likely he was to throw high in the zone.