Hello and welcome to another episode of Small Sample Size Theatre! (Motto: You know it’s highbrow because we spelled ‘theater’ wrong.) This is the show where we blow not-enough-information way out of proportion and then two weeks later after something inevitably happens to completely invalidate the article, we delete it. You won’t recall previous episodes wherein we talked about Abe Alvarez‘s vast potential, Brian Rose’s Cy Young case, and the career transformation of Julian Tavarez because we deleted them. Today’s episode: Daisuke Matsuzaka. Take a screenshot while you still can.
* * *
The Red Sox started 2-10. Remember that? Thought so. But since then? Where’s 8-1 grab ya? Me too.
In the middle of that turnaround is one of the best starting pitching stretches in the long history of the franchise, and in the middle of that stretch is every one’s favorite future Hall of Famer, Daisuke Matsuzaka. [pause for spit-take] It’s true, and it’s true that we’re only talking about four starts, which is probably a big part of what makes it true in the first place.
In fairness the entire Red Sox season has been a tale of two teams, but we’re going to focus on Matsuzaka. His four starts can easily be broken down into two start increments according to effectiveness. In terms of generalities, Matsuzaka’s first two starts were bad (mediocre and horrendous in that order). His next two were good (effective and Cy Young-caliber, respectively). The trend-line is pointing up but there is no place to go but down.
The pertinent question is, why? Why was he so awful in the first two and so effective in the next two? We can offer you at least some portion of an answer here. It should be noted that when looking at Pitch f/x data it can be dangerous to rely on pitch types as the computer doesn’t always know which pitch is being thrown. With Matsuzaka that danger is compounded because, well, does anyone really know what pitches Matsuzaka is throwing? There is some safety to be found however in what we’re about to look at.
We’re going to focus on the two most polarizing starts, the April 11th start versus Tampa and the most recent April 23rd start versus the Angels primarily because comparing those two offers the starkest contrast. The first start resulted in seven runs over two innings, the second resulted in one hit over eight innings. What was the difference in the two starts? Lets run through two possibilities.
Speed – Matsuzaka’s four seem fastball wasn’t much faster in his successful starts than in his lousy/hideous starts. As a matter of fact, his four seamer had the highest average speed in his giant mess of a start versus Tampa. That was likely a product of him only lasting two innings and trying to throw as hard as he could to finally get someone out (didn’t work). What was somewhat different was the speed of his two seamer and his cutter, both of which averaged about a mile per hour increase in speed. Verdict: Probably not the droids we’re looking for.
Pitch Differentiation – A change up is essentially a slower fastball. The good ones have movement to them, but that’s just icing on the cake. The cake itself is throwing a slow pitch that looks fast which throws off the batter’s timing. You can’t accomplish that unless there is enough of a speed change between fastball and change up. If a fastball is 91 mph like Matsuzaka’s, throwing an 85 mph change up is not slow enough to throw the batter off. Instead, it becomes just a slower fastball, one that most major leaguers can crush. Sure enough, check out the difference in speed between Matsuzaka’s change up (yellow) and his fastballs (blue) in his disastrous start versus Tampa:
The yellow dots right in the middle of the chart are change ups at about 86 mph, or four or five mph slower than Matsuzaka’s fastballs, the blue dots around 91. Now look at the same chart but for Matsuzaka’s most recent start versus the Angels:
Outside of two pitches, there’s a pretty big differentiation between fastball and change up. Keep in mind that a change up works with and sets up the fastball. As for the two graphs above, it’s pretty easy to see the difference. Against Tampa Matsuzaka’s change up and fastball were about six mph apart on average. Against the Angels that difference was about 12 mph on average. Anyone who remembers effective Keith Foulke (2004) and ineffective Keith Foulke (just about anytime after that) will recall that change ups work best when they are at least a 10 mph slower than the fastball.
The pink pitches at about the same speed as the change up are the curveball, a pitch the opposite of the change up in that it’s deception comes from it’s break, not it’s speed. The curve wasn’t enough to throw the Rays off the track, likely because they could see it coming out of his hand. When done properly the change up prevents the batter from identifying the pitch until it’s too late but without that deception in his arsenal Rays hitters could easily time his pitches and tee off. Here’s guessing that was a big part of Matsuzaka’s success or failure.