I am still kind of reeling from the outing that Daisuke Matsuzaka put forth this week in Detroit. Eight walks!!! Matsuzaka only threw 55% of his 109 pitches for strikes in what was an incredibly frustrating, yet oddly effective outing (at least in terms of net results).
With Daisuke, I never know where to turn to analyze his performance. I don’t necessarily feel like I have a clear understanding of his philosophy on the mound, nor do I know at this point which pitches he feels comfortable with in what situations. He’s got such an arsenal, it makes him hard to come to grips with. In a sense, that same arsenal must make it hard to get and keep a feel for any given pitch in any given outing.
I recently stumbled on another fun PITCHf/x tool, this time courtesy of SOSH and Brooksbaseball.net. So I decided to take a look at all the fun charts and graphs and see what they could tell me about any differences between Daisuke’s last two starts.
4/30 – 7 IP, 0 ER, 2H, 2 BB, 4 K (PITCHf/x)
5/5 – 5 IP, 1 ER, 2H, 8 BB, 1 K (PITCHf/x)
What could possibly be so different from start to start to cause such wildness from Matsuzaka?
I want to preface this all by saying, what you are seeing here is really nothing more than a first take at looking at the data and my reactions. I don’t know as I pull these graphs, exactly where this exploration will lead us and if it will even end in an answer that satisfies us. But, part of the fun of the journey is the exploration itself…so here we go.
Lets start by looking at pitch type by speed over the course of the start:
5/5 – Bad Outing
4/30 – Good Outing
OK…so you can click on the links to see the larger graphs for full glorious PITCHf/x graphing goodness. One quick point about the data, this is split by pitch type and was scored by two different people that categorize the raw data slightly differently. The bottom graph has three different versions of the fastball compared to only one on the top.
What do we see different here between the “good” and “bad” Daisuke? From my eye, it looks like “good” Daisuke had higher overall velocity and greater variation in the velocity of his pitches. “Bad” Daisuke’s speeds by pitch type seem to congregate closer in bunches.
You can see the overall results in terms of balls and strikes in the next set of graphs.
5/5 – Bad Outing
4/30 – Good Outing
In Daisuke’s “good” outing, he threw 62% of his pitches for strikes with hitters swinging at nine of the forty-two balls he threw out of the zone (21.4%). In his “bad” outing he threw strikes 55% of the time and batters swung on pitches out of the zone only 14% of the forty-nine balls.
What does that tell us? In his good outing he was around the zone more and when he went out, people chased a little more often. In his bad outing, his lack of command didn’t incent batters to chase balls out of the zone.
While there is a TON of data here about spin and movement, my head is spinning trying to make sense of it. As often happens when my head spins, I decide instead to run from what is causing that spinning and search for something to make sense of.
In this case, variances in Daisuke’s release point on his pitches caught my attention.
5/5 – Bad Outing
4/30 – Good Outing
If you look at the relative position of the clusters, there is a defanite difference in Matsuzaka’s release point and most noticeably on curve balls. His fastball(s) hover between 5 and 6 on the vertical axis and between -3 and -2 on the horizontal in both outings. However, his breaking pitches, shift between the outings. In his good outing, his breaking pitches cluster in release point with his fastballs. In his bad outing his release point is noticeably higher and to the right for most breaking pitches.
Not knowing more about where Daisuke’s natural release point is and if it should be consistent from pitch type to pitch type, all I can say is that there is a clear difference between the two outings in this regard. As you look through the spin and movement data, it looks like there are distinct differences in the ball having more downward spin away to right right handers in his bad outing.
What does this all tell us? Honestly, I am not at all sure. Believe it or not, I don’t have all the answers!
But one thing that I can say is that it will be interesting to see if patterns emerge as I expand this past these two games. My first pass through this new tool has absolutely given me reason to go ahead and learn more about what it can tell us.
For more on Paul and I’s thoughts on DiceK’s performance this week, hit up the 9:33 mark of this week’s podcast.
