Scott’s Miscellany – Matt Harvey and High-Stress Pitches

The title of the article is an allusion to Schott’s Miscellany, which you should definitely check out if you never have and feel compelled to know that a group of larks is called an exaltation or that a member of the 32nd degree of Freemasonry is known as a Sublime Prince of the Royal Secret.

—Matt Harvey and High-Stress Pitches–

Like most of the baseball world, I’m fascinated by Matt Harvey and the question of whether it is worth it for him to exceed his surgeon-prescribed innings limit because of the Mets’ place in the standings this year. To start, let me stress that I have absolutely no medical training. As a layman, I find it interesting that we rely on innings pitched as the barometer of stress for a pitcher. I expect the reason we do that is because innings pitched is the most accessible measurement of how much pitching a pitcher did. Innings pitched no doubt correlates well with total pitches thrown, so there is little lost by relying on the easier stat to find. Still, would pitches thrown even be the best way to capture the stress a pitcher faced?

I’m not the first person to ask that question. The reason I’m writing this article is because I’ve seen several quotes about Harvey—also from laypeople—that suggest that it is the stress of the pitches as much as the quantity that could threaten the health of his surgically repaired ligament. If that is the case, we can study it. To my mind, a pitcher would become more stressed when the game situation he faced became more important. Fortunately, FanGraphs already has a statistic for the importance of a game situation. It is called leverage index (LI).

To get at the answer I was looking for, I calculated leverage index independent of what you can find on the site, but the logic I used is much the same. First, I calculated the win expectancy of every state of inning, inning half (top or bottom of the inning), outs, baserunners, and score differentials based on how frequently teams in those states won those games since 2002. To help maintain sample size at the extremes, I replaced every extra inning with the ninth inning and replaced any score differentials over five runs with five runs. Then, I calculated leverage index by finding the standard deviation of the change in win expectancy from the start of at-bats to the end of at-bats and scaling that to the standard deviation of every start of an at-bat to the end of that at-bat.

Once I’d calculated the leverage index of every at-bat, I then assigned that leverage index to each pitch within the at-bat. Finally, I’m relying on the glossary recommendations of leverage indices of 2.0 or more as high-stress pitches and leverage indices below 0.85 as low-stress pitches.

In the end, Harvey was in the middle of the pack. 436 of his 2,528 pitches this year have been high-stress pitches. That is the 48th-most of the 81 qualified starters. That means that 17.2 percent of his pitches have been high-stress pitches, which is tied for the 27th-highest rate of the 81 qualified starters.

Here are the starters with the most high-stress pitches this season:

Most High-Stress Pitches, 2015

Pitcher	Pitches	High-Stress Pitches	High-Stress Rate	Low-Stress Pitches	Low-Stress Rate
Jordan Zimmermann	2833	574	20.3%	1690	59.7%
Wei-Yin Chen	2692	571	21.2%	1535	57.0%
James Shields	3107	570	18.3%	1786	57.5%
Chris Archer	3170	560	17.7%	1956	61.7%
Tyson Ross	3025	548	18.1%	1883	62.2%
Julio Teheran	2953	547	18.5%	1795	60.8%
Edinson Volquez	2928	535	18.3%	1710	58.4%
Wade Miley	2869	532	18.5%	1642	57.2%
Chris Sale	3000	529	17.6%	1760	58.7%

Here are the starters with the highest rate of high-stress pitches this season:

Highest Rate of High-Stress Pitches, 2015

Pitcher	Pitches	High-Stress Pitches	High-Stress Rate	Low-Stress Pitches	Low-Stress Rate
Wei-Yin Chen	2692	571	21.2%	1535	57.0%
Jordan Zimmermann	2833	574	20.3%	1690	59.7%
Mike Pelfrey	2439	472	19.4%	1407	57.7%
Matt Garza	2378	442	18.6%	1383	58.2%
Wade Miley	2869	532	18.5%	1642	57.2%
Julio Teheran	2953	547	18.5%	1795	60.8%
John Danks	2599	477	18.4%	1633	62.8%
Edinson Volquez	2928	535	18.3%	1710	58.4%
Jon Lester	2866	525	18.3%	1651	57.6%

And here are the starters with the lowest rate of high-stress pitches this season:

Lowest Rate of High-Stress Pitches, 2015

Pitcher	Pitches	High-Stress Pitches	High-Stress Rate	Low-Stress Pitches	Low-Stress Rate
Jake Arrieta	3036	377	12.4%	2063	68.0%
Marco Estrada	2510	322	12.8%	1639	65.3%
Anibal Sanchez	2523	337	13.4%	1723	68.3%
Carlos Carrasco	2487	340	13.7%	1568	63.0%
Mike Fiers	2725	379	13.9%	1696	62.2%
Alex Wood	2576	363	14.1%	1516	58.9%
Danny Salazar	2743	388	14.1%	1802	65.7%
Mark Buehrle	2500	365	14.6%	1594	63.8%
Jacob deGrom	2803	418	14.9%	1761	62.8%