Breaking Down Contact Rate by Count

I’ve been in love with contact rate from the beginning. Admittedly, it doesn’t provide a lot of information you can’t already figure out from strikeout rate, but I like contact rate for its granularity. It’s one of the first stats I look at for pitchers, and it seems like a pretty pure indicator of domination over opponents. What could be better, after all, than a pitch a batter can’t hit? An out on a ball in play, I suppose, but balls in play are dangerous. Nothing dangerous about a swing and miss.

As much as I like contact rate, though, I’ve never thought to try to break it down. Most analysis with contact rate is performed using overall contact rate. Sometimes, it’s using in-zone contact rate, or out-of-zone contact rate. But what I found myself in the mood to do is try to break the numbers down a little bit by count. A thank you, as usual, is extended to Baseball Savant. Because I haven’t spent much time with this data, I can’t really tell you what it means yet, but, you’re here because you want to further your understanding of baseball. I think exploration’s justifiable for exploration’s sake.

I don’t know how many times I’ll play with this data — we’ll see where this might lead, if anywhere — but for this first pass, I’ve opted to keep things simple. Instead of overall contact rate, I looked for pitcher contact rate allowed with the pitcher ahead, and I looked for pitcher contact rate allowed with the pitcher not ahead. Those counts considered for the former: 0-and-1, 0-and-2, 1-and-2, 2-and-2, and 3-and-2. The last one was a judgment call, but I basically wanted to capture all two-strike counts, and also 0-and-1.

I’ve used only 2014 data, and I’ve established a minimum of 200 swings. Let’s start with some tables.

Lowest contact rates allowed, pitcher ahead

Rank	Pitcher	Contact%
1	Aroldis Chapman	54.2%
2	Andrew Miller	57.1%
3	Brett Cecil	57.7%
4	Greg Holland	58.6%
5	Will Smith	60.3%
6	Craig Kimbrel	60.8%
7	Joaquin Benoit	61.4%
8	Tony Sipp	61.9%
9	Dellin Betances	62.3%
10	Jordan Walden	62.7%

You’re not surprised! You shouldn’t have expected to be surprised. When Aroldis Chapman got ahead, his contact rate came in incredibly close to 50%. He’s not easy to hit at the best of times, but when you cede control of the count to Chapman, the situation becomes even more unfair. Andrew Miller is also elite. Brett Cecil is better than you might have realized. And so on and so forth. Tony Sipp was a good pick-up by the Astros after he got dropped by the Padres.

Highest contact rates allowed, pitcher ahead

Rank	Pitcher	Contact%
1	Burke Badenhop	88.5%
2	Anthony Swarzak	87.2%
3	Mark Buehrle	86.7%
4	David Huff	86.6%
5	Nick Tepesch	85.5%
6	Doug Fister	85.2%
7	Vance Worley	85.1%
8	Christian Bergman	84.5%
9	Scott Carroll	84.5%
10	Bruce Chen	84.4%

These aren’t all bad pitchers — they’re just pitchers who don’t really have obvious putaway pitches. Vance Worley likes to nibble around the edges, expanding the zone as much as he can with his fastballs. Mark Buehrle tries to induce weak contact when ahead. The same goes for Doug Fister. On the one hand, we know better than to believe in weak contact as a strongly sustainable skill; on the other hand, there’s a gray area. If you’re wondering who Christian Bergman is, you’re not alone, and this is my professional job.

Lowest contact rates allowed, pitcher not ahead

Rank	Pitcher	Contact%
1	Koji Uehara	61.5%
2	Tyler Clippard	62.0%
3	Wade Davis	62.4%
4	Randall Delgado	62.8%
5	Alex Torres	63.3%
6	Craig Kimbrel	64.7%
7	A.J. Ramos	65.0%
8	Francisco Liriano	65.3%
9	Mike Dunn	65.7%
10	Antonio Bastardo	68.0%

A problem with setting a 200-swing minimum and looking at contact rate is that you invite a lot of relievers to the top of the lists. But here, Francisco Liriano sneaks into the top-ten, as a guy who wasn’t easy to hit even when he didn’t have the count in his favor. Liriano narrowly missed showing up in the previous low-contact top-ten. As frustrating as he can be some of the time, he never makes it easy to put the bat on the ball, which is an enviable skill. As far as Koji Uehara is concerned, when you think of him you think of his putaway splitter, but he doesn’t limit throwing that to putaway counts. Tyler Clippard throws a ton of changeups when he’s not ahead. Wade Davis is always willing to throw a curveball or a cutter. And also, he’s just Wade Davis.

Highest contact rates allowed, pitcher not ahead

Rank	Pitcher	Contact%
1	Joe Saunders	88.7%
2	James Paxton	88.1%
3	Kevin Correia	87.1%
4	Yohan Pino	86.2%
5	Tyler Matzek	86.2%
6	Scott Carroll	86.0%
7	Paul Maholm	85.7%
8	David Phelps	85.6%
9	Ross Detwiler	85.5%
10	Brandon League	85.4%

I think there are two surprises here — James Paxton and Tyler Matzek. The rest of these guys are generally hittable. Paxton is a highly-prized Mariners asset. Matzek is a little less highly-prized, but the Rockies do value him. Conveniently, we’ll touch on these guys more in a minute.

You might be wondering how the two contact rates relate to one another. This is how they do that, out of a sample of 280 pitchers:

Obviously, there’s a direct, linear relationship. Yet there are points that stray pretty far from the best-fit line, and people are always interested in the outliers. I calculated z-scores for the two different contact rates, then I calculated the difference between them. Let’s now look at the ten most extreme pitchers in both directions.

Greatest difference in contact rates, favoring being ahead

Rank	Pitcher	Contact%, not ahead	Contact, ahead	z, not ahead	z, ahead	Difference
1	James Paxton	88.1%	71.8%	-2.3	0.9	-3.1
2	Will Smith	75.0%	60.3%	0.4	3.1	-2.7
3	Tyler Matzek	86.2%	73.0%	-1.9	0.6	-2.5
4	Yu Darvish	79.9%	66.9%	-0.6	1.8	-2.4
5	Dellin Betances	74.5%	62.3%	0.5	2.7	-2.2
6	Jacob Turner	84.3%	73.9%	-1.5	0.4	-1.9
7	Tony Cingrani	81.9%	72.3%	-1.0	0.8	-1.8
8	Junichi Tazawa	78.5%	69.2%	-0.3	1.4	-1.7
9	Francisco Rodriguez	76.2%	67.2%	0.2	1.8	-1.6
10	Fernando Rodney	78.2%	69.4%	-0.2	1.3	-1.6

Here, again, we see some of the names from above. Leading the way is James Paxton. When not ahead, his contact rate was more than two standard deviations worse than the mean. When ahead, it was almost one standard deviation better than the mean. So you could interpret this in one of two ways: either Paxton is excellent when ahead, or he struggles when he’s not ahead. What’s the likely explanation for this? Paxton’s known for his devastating curveball. When ahead, he threw about 35% curveballs. When not ahead, he threw about 7% curveballs. Though Paxton throws hard, his fastball isn’t tough to hit. Paxton’s curve is pretty clearly a swing-and-miss pitch, not a pitch he trusts to throw for a strike.

Will Smith threw about 56% sliders when ahead. That fell to 24% when not ahead. Tyler Matzek also threw breaking balls about half the time when ahead. When not ahead, he was roughly three-quarters fastballs. Yu Darvish last year threw more fastballs in general, perhaps in an effort to get quicker outs and put less stress on his elbow, but even still, he threw about 50% secondary stuff when ahead. When you fall behind Darvish, you don’t know what’s coming, and he’s got an array of weapons. When Darvish isn’t ahead, he might throw about three-quarters fastballs, up from the previous year’s 64%.

Greatest difference in contact rates, favoring not being ahead

Rank	Pitcher	Contact%, not ahead	Contact, ahead	z, not ahead	z, ahead	Difference
1	Alex Torres	63.3%	75.7%	2.8	0.1	2.7
2	Tyler Clippard	62.0%	72.9%	3.1	0.7	2.4
3	Randall Delgado	62.8%	72.0%	2.9	0.8	2.1
4	Brian Matusz	71.6%	80.4%	1.1	-0.8	2.0
5	Jonathan Papelbon	70.0%	77.6%	1.4	-0.3	1.7
6	Erik Bedard	75.1%	82.9%	0.4	-1.3	1.7
7	Allen Webster	70.1%	77.2%	1.4	-0.2	1.6
8	Hector Santiago	76.9%	83.7%	0.0	-1.5	1.5
9	David Huff	79.7%	86.6%	-0.5	-2.0	1.5
10	Robbie Erlin	75.5%	82.1%	0.3	-1.2	1.5

Now it’s Alex Torres’ turn. When ahead, or when not ahead, Torres threw basically identical rates of changeups. Unlike James Paxton, who wants to throw his curveball out of the zone, Torres likes his changeup as a pitch he can throw for a strike, and as a pitch he can throw when the hitter is looking for something straighter and faster. Tyler Clippard throws fewer changeups when he’s ahead, as he likes to try to climb the ladder with his heat. Randall Delgado and Brian Matusz, like Torres, also keep their pitch rates similar. Delgado, when not ahead, had good success with a low arm-side changeup. This is a list with some changeups. Might be something about changeups.

What I don’t have is a big conclusion right now. This was strictly exploratory, and I don’t know where this might go next. As far as whether it’s better to have a superior contact rate when ahead or when not ahead, it appears, based on 2014 information, that it’s a little better to have a superior contact rate when ahead. That is, if you have to have a big difference between your contact rates, you’d rather have a lower contact rate when in control of the count. But, subjectively, the relationship isn’t very strong. Miss bats when ahead, and you’re directly putting guys away. Miss bats when not ahead, and you’re working out of pitcher-unfriendly counts. So there are advantages and other advantages.

More than anything else, you just want to have a low contact rate in general. That much, we already figured. But there’s a whole lot of ways to slice this data up.