Today I wanted to take a look at some data and see if we can make any sense of any changes that MIke Maddux has actually (not theoretically) brought to the the St. Louis Cardinals in his two years as the pitching coach, when he replaced former coach Derek Lilliquist after the 2017/prior to the 2018 season. (At this point I'd like to point out that I was nearly 10,000 characters, 8 .gifs, and 2 .jpgs into this article before it was brought to my attention that @stlcupofjoe wrote about this for The Athletic back in 2018 as well. I think this will stand on it's own, but as always Joe did a fantastic job and if you have a subscription to The Athletic, you should read it as well.)
First, let's take a look at the pitchers that both pitching coaches had to work with in the two seasons prior to and post-coach switch. The Derek Lilliquist led 2016-17 group is on the left and the Mike Maddux led 2018-19 group is on the right. Italicized are the players that played for BOTH pitching coaches.
In looking at that group of players, I felt it would be unfair to compare what the players that did not pitch for both pitchers did while a Cardinal. My thought is wouldn't it be more fair to compare the same groupings of data? So I only looked at the following players' statistics for the remainder of this article - the group of 14 below (alphabetically):
Matt Bowman
John Brebbia
Brett Cecil
Jack Flaherty
John Gant
Tyler Lyons
Carlos Martinez
Mike Mayers
Alex Reyes
Ryan Sherriff
Sam Tuivailala
Michael Wacha
Adam Wainwright, and
Luke Weaver
There was nobody else who pitched for both Lilliquist in 2016-17 and Maddux in 2018-19. Did those 14 pitchers change the usage of their repertoires after the coaching change? First things first, let's do some large groupings here. The group as a whole threw 1,564 2/3 innings under Maddux and 1,630 1/3 innings under Lilliquist in the last two years of his time here. Those are fairly equal - with Maddux's tenure seeing a reduction of just 4% of the innings is all.
STEP 1 - Broad pitch categories - SIMILAR PITCHERS
I want to start by taking a look at 6 gifs. There will be two about fastballs, one on location and one on K% by area of the zone. There will be two about offspeed pitches - location and K%. There will be two about breaking pitches - location and K%. I will add a small analysis of what I see in those gifs that you can determine, as the reader, if you agree . Here they are:
Change in Fastball Location
Fastballs have decidedly been thrown higher in the zone and higher out of the zone under Maddux than under Lilliquist. This was one observation that I came into this article expecting to find. It was something that Maddux preached upon coming over - more fastballs up in the zone. Lilliquist was more of a sinkers down guy - a Dave Duncan protegee, if you will.
Change in Fastball K%
While fastball strikeout totals have stayed similar at the bottom of the zone for Cardinals pitchers through the pitching regime change, the percentage of strikeouts in the zone and up in the zone on fastballs has risen precipitously.
Change in Offspeed Location
The change in offspeed location has been to lower it in the zone slightly, it seems. There' s not a huge change that I see here, philosophically, unlike the fastballs above. This is a spot where it seems Lilli and Maddux agree completely.
Change in Offspeed K%
Cardinals pitchers seem to be getting similar strikeout totals on offspeed pitches (percentage-wise) under both Lilliquist and Maddux, but where they are getting them is different. I'll let you look and see on this one because a picture (or .gif) is worth a thousand words and I don't know that I could explain this better than the ol' "eye test" on this one...
Change in Breaking Ball Location
This data actually is counter-intuitive to what I thought entering the article. Under Maddux, the zone has shifted even further down on breaking pitches, something I expected to see slightly differently. Just going by my (apparently elusive) memory, I would have sworn over the past two years to have seen many more breaking balls near the top of the zone playing off of the fastballs up. I guess I'm wrong - unless it's the pitchers not listed that are doing so. In any case, it's clear that more breaking pitches by this group were thrown down than before - and not by an insignificant margin.
Change in Breaking Ball K%
It seems as though the change in location of the breaking pitches has not helped so much the strikeout percentage on breaking balls by the members of the Cardinals' staff listed above. In fact, the gain made by fastballs in the strikeout department were probably the direct result of being taken away from the breaking ball strikeouts from the previous regime.
That last breakthrough (the percent of breaking balls going for strikeouts basically morphing into fastballs going for strikeouts) leads me to this; In the change over from Lilliquist to Maddux, there was a 10 percent increase in frequency of throwing two-strike fastballs and a 5 percent decrease in frequency of throwing two-strike breaking pitches. I think that would do it.
In other words, we've seen more of these lately:
HOWEVER, when Lilliquist was here we saw a bit more of this:
STEP 2 - Individual pitch categories - SIMILAR PITCHERS
What else can we find if we look at percent of pitches thrown? Here are all of the different pitch types thrown by these 14 pitchers as Cardinals broken up by "Under Lilliquist" or "Under Maddux." The last column is simply the change in percentage - the deeper the red means the more that pitch was thrown while the deeper the blue means the less that pitch was thrown - and remember, this is by the same pitches.
The obvious stand outs are the four-seamer and slider gaining the two highest percentages while the two-seamer (remember, the Lilliquist staple) losing the most and the changeup/cutter both losing some as well. The splitter and "other" numbers basically don't matter...while the curve/knuckle curve were basically thrown with the exact same frequency as before.
If fastballs are thrown higher in the zone under Maddux, as we saw, then it makes sense to grow the difference between four-seamer and two-seamer from a 12% difference under Lilliquist to a 20% difference under Maddux. My theory that there were simply more four-seamers thus more fastballs thrown up in the zone is incorrect, however. Take a look at the percentages of four-seam fastballs by location in 2016-17 (12.2 and 11.2 percents up out of the zone) compared to 2018-19 (16.5 and 12.6 percents up out of the zone) this .gif fades from the former to the latter.
As is clearly shown, the four-seamer has risen in the zone significantly during Maddux's time as pitching coach. Let's look at some individual pitchers to see how this has affected them. This is where things get really interesting to me - and where this article will take a bit of a turn towards what I see as an obvious truth - at least to an outside observer. It's not what I was expecting.
Adam Wainwright, Michael Wacha, and Carlos Martinez threw some of the highest inning totals under both pitching coaches in the 2016-2019 era. We'll look at those three. Here is the heatmap for Adam Wainwirght's four-seam fastball fading from the Lilliquist era first to the Maddux era second. Prior to looking at this, one thing to note is that all prior strike zone location images (above, from Baseball Savant) have been from the catcher's perspective whereas these heat maps (from Fangraphs) are from the pitcher's perspective.
In 2016-17, under Lilliquist, Adam Wainwright's 4-seam fastball was thrown all over the strike zone. Notice how Adam Wainwright's fastball really changed. It was focused IN on a right-handed hitter or away from a left-handed hitter almost exclusively the last two seasons, under Maddux. While it didn't rise, necessarily, it was very concentrated.
Even more fascinating to me is that we see the same phenomenon occur with Michael Wacha! Look at this:
Prior to looking at Carlos Martinez's, my thoughts are that Maddux is just much better at honing in on consistency for pitchers than Lilliquist was - whether that be mechanics or gameplan or the like.
Let's look at what Martinez's four-seam shift can show us:
There is not a clear shift for the majority (darker portions) of the data - potentially slightly upward if anything. But if you look at the other concentrations (yellow and lighter orange colors), it seems to me that the general trend is less concentrated closer to the zone - alternatively, less consistent - with a slight shift upwards in (and out) of the strike zone as well. A couple of theories can fill the void here, but not necessarily be correct. Maybe Martinez was throwing harder and more wildly out of the pen in recent stints? Maybe Martinez's injury in 2018 really threw off four-seam fastball command? Who knows.
In any case...with those three main pitchers that logged a lot of innings, the upward trend was actually not consistently there...which was a surprise to me. The consistency stuck out and it seemed as though there was a shift IN to a righty (away from a lefty).
In looking at the two-seam / sinker, there was really no change in that group of 14's usage by location. They just threw less two-seamers than before to the same locations - as a collective. This group threw nearly the exact same amount of curves, but breaking balls as a whole were thrown more often and the zone thrown was shifted drastically downward. Maybe the slider is where we look next.
HERE is the clear change. When you look at the next GIF, look first at the key to the right and notice that in 2016-17, the Cardinals pitchers we are looking at did not throw nearly as many sliders to left-handed hitters (LHH) as they did the last two years. Maybe that's because they were hanging them barely on the outside corner for the most part rather than burying them lower in the zone to induce grounders more often, as they did in 2018-19 with Maddux.
When we look at the same type of GIF, but on sliders to right-handed hitters (RHH), consider the word we used above; consistency. Under Maddux in 2018-19, not only are the sliders to lefties and righties quite consistent, but the sliders solely to RHH are ridiculously consistent in their location.
The more I look into Mike Maddux's time in St. Louis the last two seasons, the more that I find consistency could very well be the key that Maddux is trying to use to unlock the potential of the St. Louis Cardinals pitching staffs year-to-year.
Next, let's look at slider usage with some of the individual pitchers from the past 4 seasons in Cardinals' history. Out of the 14 pitchers above, the 2 pitchers that threw the most sliders were John Brebbia and Carlos Martinez - with Tyler Lyons also throwing quite a few. Let's look at how those 3 changed usages of sliders from 2016-17 to 2018-19.
Here is Carlos Martinez's slider usage under both coaches:
In both images, the larger clump near the middle of the zone is his slider vs. left-handed hitters. Those images really didn't change all that much. However, what did change was that against right-handeed hitters, his location got much better, lower in the corner of the zone and just out of the zone. Much less hittable pitches down there against righties - at least more consistently.
So did John Brebbia - who has lived on the fastballl-slider combo - change his lcoations of his sliders greatly? This time, do me a favor. At this point we can probably guess without looking at the dates below the images which year is which. Try that here - cover it up with your finger if you need to.
You likely guessed it correct. In 2018-19, Brebbbia's slider just seemed so much more consistent via that heat map didn't it? Wow. That is definitely the unexpected theme of this article.
On to Tyler Lyons:
OMG. Did Ben label those wrong? The answer to that is no. So my whole last 5 minute soliloquy on consistency gets blown up with this one. Oof.
In any case, so far I think the overwhelming success story here is Mike Maddux imparting consistency on his pitchers. I truly do. Tyler Lyons' slider usage did not show consistency like the others - that doesn't mean it wasn't Maddux's message. As you notice, however, when Lyons' slider usage was more consistent you'll notice that he had much better results.
STEP 3 - UNCOMMON PITCHERS
The last thing I want to look at today is, have the Cardinals brought in pitchers that they believe espouse Maddux's values we have found (fastballs up (and in) in the zone) and bountiful sliders (consistently low in the zone) - compared to what pitchers they had here with Lilliquist in charge? In order to do so, we'll have to take a look at the pitchers the Cardinals have gotten rid of vs. the pitchers the Cardinals have brought in.
Here are the 16 pitchers that the Cardinals have lost, that pitched in 2016-17 under Lilliquist: Mike Leake, Lance Lynn, Jaime Garcia, Seunghwan Oh, Kevin Siegrist, Trevor Rosenthal, Jonathan Broxton, Zach Duke, Miguel Socolovich, Seth Maness, Dean Kiekhefer, Jerome Williams, Juan Nicasio, Sandy Alcantara, Josh Lucas, and Marco Gonzales.
Here are the 19 pitchers that the Cardinals have gained, that pitched in 2018-19 under Maddux: Miles Mikolas, Dakota Hudson, Jordan Hicks, Daniel Ponce de Leon, Giovanny Gallegos, Austin Gomber, Tyler Webb, Dominic Leone, Bud Norris, Andrew Miller, Ryan Helsley, Tyson Ross, Greg Holland, Genesis Cabrera, Luke Gregerson, Chasen Shreve, Junior Fernandez, Adalberto Mejia, and Preston Guilmet.
Let's start by just looking to see if the group that have pitched here the last two years are - as a whole - a better fit under Maddux's style. The group of 16 above that Lilliquist had from 2016-17 that were NOT KEPT for 2018-19 threw 34.4% four-seam fastballs, 37.1% two-seamers/sinkers AND cutters combined, 4.9% curves/knuckle curves, and 11.5% sliders. The group of 19 above that Maddux had from 2018-19 that were ADDITIONS TO the major league club after Lilliquist left threw 31.9% four-seam fastballs, 29.4% two-seamers/sinkers AND cutters combined, 10.6% curve balls, and 22.4% sliders.
Analyzing those numbers, while the pure percentage of four-seamers went down, it was 52% of all fastballs thrown compared to just 48% of fastballs thrown under Lilliquist. Furthermore, you can see how many more breaking balls the guys the Cardinals brought in for Maddux threw, on the whole! It's alarming. Those new guys' breaking ball usage was 201% the usage of the guys let go!!! Not only that but look at the slider usage. It went from just over 1 of every 9 pitches to just over 2 of every 9 pitches thrown! WHOA.
I'd say that, in fact, the 19 pitchers that Maddux has had that were not in the majors under Lilliquist definitively have looked different on the whole from purely an arsenal standpoint. I will spare you the details and let you know that the fastballs not only were a much higher percent of four-seamers, they were placed WELL UP in the zone in comparison. Honestly, the location of the slider didn't really change at all, but was just doubled in how often it was deployed.
Another interesting happenstance is how often this group (unlike the ones talked about in Steps 1 and 2) has gone back to high curve balls, despite the lower sliders. In Joe's piece I linked to from The Athletic, he talked about how Miles Mikolas had stolen many first pitch strikes on high curve balls. Well, over the last two years, he has led the way for the Cardinals with 9 strikeouts on called third strikes on curves up and in/out of the zone. (Here they are below.) The Cardinals had over 30 of those and were 6th in the majors in such strikeouts.
OK. I think I have utilized enough time of yours today. Make any conjectures you can from this and get back to me on Twitter.
I'd like to thank Baseball Savant, Fangraphs, and @cardinalsgifs for their fantastic help in getting this article out there for all of you wonderful readers!