Carlos Marmol, expanded

Wow. It's been quite a while since I posted a player analysis piece. Just been waaaay too busy getting ready for law school I guess, and now that I'm employed it's hard to find time to piss away reading player charts daily. Aaaanywho, in case you missed it, I was on 670 The Score earlier today talking about Marmol's struggles. My assertion was, is, that Marmol's struggles are not related to mechanics but adjustments by major league hitting (specifically in reference to a drastic reduction in swing rates against his slider). Can I prove it? Read on to find out!

Take a look at Marmol's three year Pitch FX data, courtesy of Fangraphs.com (click the image to enlarge):



There are several things to note about this chart of data:
1) Marmol's velocity on each pitch is right in line with his three year averages (or slightly higher) -- IN-Vel is irrelevant because it refers to intentional balls.
2) Marmol's horizontal slider movement is on a three year decline, but still within 0.3 inches of his 2008 horizontal movement levels.
3) Marmol's vertical movement on the slider has been consistent each of the past three seasons.
4) Marmol's fastball movement (vertical and horizontal) and velocity has been consistent during his career as a reliever.
5) Marmol's curveball has maintained its vertical movement each season, but lost some horizontal movement each of the past three seasons -- although his 2009 horizontal movement is within a 0.4 inches of his 2008 movement level.
6) Marmol's curveball has gotten faster each of the past three seasons.

The previous six points reiterate and reinforce my thesis: that Marmol's stuff has been as good as its ever been. Especially compared to last year, Marmol's break on each pitch has been the same and his velocities have also the same (or better) on each of his pitches. So why, if his stuff has been the same, has Marmol's runs prevention per 100 pitches value on the slider dropped from a peak of 21.4 in 2008 to a career low 3.8 in 2009?

In my opinion, the answer is not that he's been overworked each of the past two seasons (he leads all relievers in IP from 2007-2008). The true answer lies in the following two charts (again, click to enlarge):




Each of the last two seasons, hitters have increasingly come to recognize that Marmol's sick slider usually breaks outside of the zone when thrown low -- and thus can be taken for a ball rather than swinging strike. Since his 2007, Marmol has induced about 5% less swings (per season) on pitches thrown outside of the zone. Furthermore, as hitters have been swinging at less and less of his pitches in general, his walk rate has skyrocketed from 4.54 per nine innings in 2007 to a ridiculous 8.48 rate per nine in 2009, while the strikeout per nine rate has declined from 12.46 to 10.56 over that same time frame. What is surely doing most of this damage is an increase reliance by Marmol on his slider. Since 2007, Marmol's slider usage has increased from 31.5% to 38.2% to a current rate of 40.0%. Thus, while Marmol's slider has been equally effective in terms of pure stuff each of the past three years, hitters have caught on to the fact that the low slider will break for a ball and thus can be taken for ball four rather than a swinging strike three.

The only remedy, as Boers and Bernstein indicated in their reply to my call, is an adjustment in Marmol's pitch selection and location. Unless he starts mixing up his pitches more effectively -- working in more, better controlled fastballs, for instance -- and holding off the low slider, Marmol's effectiveness as a top stop reliever will continue to decline.