Matt Harrison: The Good, The Bad & The Ugly

Texas Rangers left-hander Matt Harrison has been two different pitchers this season. One of them is a promising #3 starter who can rattle off 20-plus-inning shutout streaks. The other is a pitcher who could use some time in AAA to work things out.

To the naked eye, nothing really seems different between these two incarnations of Harrison. Thanks to PITCHf/x data, we don't have to use our naked eyes to find out what's different.

I've broken Harrison's season up into three pieces: the good, the bad, and the ugly. The good is Harrison's three-start stretch from May 3 through May 14. The bad is his first three starts of the season, April 11 through April 22. The ugly is his most recent two starts, May 20 and May 25.

Here's a quick breakdown:

- - -	IP	H	R	ER	BB	K	HR	ERA	WHIP
Good	23.0	15	2	2	1	15	0	0.78	0.70
Bad	14.2	26	15	15	11	6	4	9.20	2.52
Ugly	10.0	20	12	12	3	5	3	10.80	2.30

There isn't a ton of difference between the bad and the ugly, but in looking at the PITCHf/x data, there's reason to separate the two. The following chart breaks down Harrison's pitch selection, velocity and strike percentages for each pitch:

- - -	Bad			Good			Ugly
Pitch Type	Sel. %	Vel.	Strike %	Sel. %	Vel.	Strike %	Sel. %	Vel.	Strike %
FF	48.23	90.36	59.56	65.06	91.18	67.49	54.21	91.01	68.93
FT	16.31	87.12	54.35	9.94	86.55	83.87	25.26	86.23	45.83
CH	13.83	82.65	56.41	13.46	81.85	50.00	6.32	84.98	25.00
SL	9.22	83.98	38.46	6.09	84.83	36.84	6.84	84.81	84.62
CU	7.80	76.19	54.55	2.56	73.61	75.00	5.26	74.13	60.00

PITCH SELECTION

One thing that jumps out right away is Harrison's use of the fastball. Bad Harrison threw his four-seam fastball 48.23% of the time and his two-seam fastball 16.31% of the time. Bad Harrison's left/right splits did not show much variation.

Good Harrison's pitch selection was quite different. He threw his four-seam fastball 65.06% of the time, while only throwing his two-seam fastball 9.94% of the time. Against lefties, he threw four-seam fastballs nearly 80% of the time mostly at the expense of his two-seamer. His right-handed splits were basically in line with his overall numbers.

Ugly Harrison threw his two-seam fastball nearly three times more often than Good Harrison, and threw his four-seam fastball and change up less often as a result.

THROWING STRIKES

You can tell simply by looking at the standard stat line that Bad Harrison's control was pretty bad. He didn't reach the 60% strike percentage barrier on any one of his 5 pitches.

Good Harrison threw his four-seam fastball for a strike nearly 70% of the time, and even though he was throwing his two-seam fastball less often, he clearly had beter command of it. Good Harrison's strike rate was 83.87% for his two-seam fastball. For Bad Harrison, it was a strike less than 55% of the time.

Ugly Harrison, while throwing two-seam fastballs more often, managed to throw the pitch for fewer strikes. His strike rate plummeted to an abysmal 45.83%.

Ugly Harrison saw a huge spike in his ability to throw his slider for strikes. This is an anomoly, though, and not an indication of throwing "too many" strikes. It was only put in play 4 times - twice for outs, twice for singles.

VELOCITY CHANGE?

If you look at the velocity averages from PITCHf/x, you'll see that Ugly Harrison's two-seam fastball and change-up have a much smaller velocity separation than when thrown by either Bad Harrison or Good Harrison. This could just be noise caused by the PITCHf/x pitch type algorithm, or it could indicate that Harrison has been struggling with his change up in his last two appearances.

It should be noted that even when there was a 4-5 MPH separation between the two pitches, the spin angle and spin rate of the two pitches were nearly the same, and the algorithm was able to separate them fairly well.

If the PITCHf/x data is correct, Ugly Harrison's change up and two-seam fastball are practically the same pitch, and it's usually not a strike.

PUTTING IT ALL TOGETHER

Good Harrison relies heavily on his four-seam fastball, and he throws it for strikes.

Bad Harrison struggled to throw strikes and may not have been throwing four-seam fastball often enough.

Ugly Harrison likes to throw his two-seam fastball, even though he has trouble finding the strike zone with it. Ugly Harrison has also seen a small spike in the velocity of his change up.

If Harrison can fix his change-up and re-establish his four-seam fastball, then Texas Rangers fans will likely see Good Harrison again.

Comments

[Jeremy H.]

Welcome Trip, nice read.

I agree, when Matt pounds the zone he's really effective. In the beginning of the year he would nibble at corners and fall behind in counts. If he gets ahead he's a great pitcher to have. I hope his last start was just a bump in the road.

[Pabloesque]

Good Lord... where did you come up with all this data? Incredible stuff... especially when you consider the fact I hate stats like this ...
I think this verifies the fact that without a good change up... and without the ability to locate your pitches... you won't be a very successful MLB pitcher. That much is obvious.
He probably feels like he has more command of his 2 seamer and can hit his spots with it... yet he's not. The 4 seam fast ball tends to move more than the 2 seamer... so maybe that's why he's having good results with it, yet he feels like he can't control it as well (even though your data shows he has controlled it better).
Anyway... what do I know!
This is good stuff... I'd love to see a breakdown like this on Holland!

[Pabloesque]

1 other question/comment; why the velocity dip between 2 seam and 4 seam fastball? We're talking 3 - 5 MPH... is he taking something off his 2 seamer... or is it a lower velocity pitch by nature?
I always thought the 2 and 4 seam fastballs were thrown at the same velocity... the only difference being control and movement, respectively.

The other interesting thing here is that his change up kind of sucks. He needs to throw it consistently in the 80 MPH range or not throw it at all.
Not important he can throw the change for strikes... he needs to bury it in the dirt anyway... especially if he uses it after a 91-93 mph fastball.

He's still a kid... still learning HOW to pitch... he's shown flashes of brilliance, enough to prove he belongs here and can be a good #3 for years to come. Besides, he's left handed!

[rob m]

Thanks Trip. It should be noted that in Harry;s Ugly starts he had trouble keeping his FF down in the zone thus forcing him to throw other pitches.

[ncrangerfan]

Harrison is really bad when he doesn't pitch off his fastball. In his last game, every fastball he threw resulted in a hit hard. He lost confidence in the pitch and tried to get cute with his secondary stuff. It didn't work.

[Trip Somers]

Pabloesque: The two-seamer is kind of a tricky pitch. It can be faster than a four-seamer, but most are a couple miles per hour slower. The pitcher's intent makes a difference (is his goal sinking action? tailing action?), but mechanics and finger pressure play a large role as well.

The data is from MLB's Gameday system. It is publicly accessible, and I've built my own database based on it.

[Tucson Flash]

Trip; Reading your stuff gives me a headache. How can you come up with info like this...and be so bad at Trivial Pursuit!

[NCRF]

Nice breakdown.

There seems to be a correlation there, but I don't know that I'd attribute the different results to the pitch selection. It might be the other way around. If he's got runners on base, he's more likely to throw his offspeed stuff, specifically the 2-seamer, in order order to induce a double-play grounder.

I think strike % is the more telling stat. Except for Ugly Harrison's slider, Good Harrison has equal or better strike % for all his pitches than either the Bad or Ulgy versions. And Ugly Harrison's strike % for his slider can easily be written off as SSS variance since its only two starts and 13 pitches.

[Trip Somers]

Tucson Flash: I won the first game, as I recall. I think this analysis had too many numbers and not enough words in it. I almost gave myself a headache while proof-reading it.

NCRF: I called Ugly Harrison's strike % for his slider an anomoly. I left out the words "small sample size" but figured it was understood. Strike % certainly plays a role, but the thing that sets Ugly Harrison apart is that he threw more two-seamers than ever *and* that he wasn't throwing it for a strike. His four-seam strike % was still close to 70%, but he was throwing it less often and falling behind hitters as a result.