How Matt Holliday changed in 2015

Matt Holliday as you know, has received the biggest contract in Cardinals' history: $120M over seven years. And as you also know, the contract has been a smashing success as far as seven year free agent contracts for 30-year-olds go. In fact, according Fangraphs' value tab, Holliday has been worth $165M over the first six years of the contract, and that's even after a shortened 2015 that saw him tabulate just under a win of value.

So no matter what happens with Matt in his last guaranteed year, the contract will be a success. But suddenly, after losing Heyward, and trading Jon Jay and giving away Peter Bourjos (who both would have been gone after this year anyway), and a bare free agent market next offseason, the Cardinals suddenly have a vested interest in Holliday holding up enough to be worth the price of his $17M option ($16M over the $1M buy-out) following the end of this season.

Matt Holliday had a rough 2016. It started looking like it was going to be another great year for the ageless slugger. Through June 8th, Matt had a 134 wRC+. Not bad for a 35-year-old. Then disaster struck. On June 8th while chasing after a shallow fly ball, Holliday's cleat got stuck for a split second which managed to cause a grade 2 quadriceps strain.

Somewhat out of character, the Cardinals managed to make things worse. Holliday returned to the lineup on July 17th, but wasn't full rehabbed. The Cardinals had him under strict orders not to run at 100%. Well, that is probably easier said than done. Matt Holliday has been playing the game for most of his life, and it probably isn't very natural for any player to suddenly give less effort, especially in the middle of a close division race for a historically strong division. On July 29th, less than two weeks after his ill-advised return, he re-injured the same quad.

Holliday would return again late in the season, but it was still mostly a lost season for him. His 277 PA was easily the shortest season of his career. The season with the second fewest PA was his rookie year all the way back in 2004 when he accumulated 439 PA. In that short amount of time, he began some disturbing trends that should be watched in 2016.

Let's start with the batted ball data. Here's last year's stats plus the previous three years:

One eye-popping stat: A career low Pull%. Second lowest for his career was the 2012 figure shown here, so this wasn't an outlier by just a few percentage points. Going into 2015, Matt pulled 35.1% of balls in play, which has been below league average but not uncomfortably so. Players don't have to pull the ball a lot to be successful. In fact most pure hitters spray the ball all over the field, and Holliday has been no exception throughout his career. But that's not what Holliday was doing in 2015.

In 2015, Holliday had the the third lowest Pull% among players with 200 or more PA. The rest of the top 10 does not inspire confidence: DJ LeMahieu, Jean Segura, Shane Peterson, Howie Kendrick, Joey Butler, Juan Lagares, Jason Bourgeois, Ichiro Suzuki, and Casey McGehee. Of those nine players, only Joey Butler and Howie Kendrick had above average seasons at the plate, both posting a 109 wRC+. So it doesn't seem like a very good thing to have an extremely low Pull%. And it seems hard to be even above average with such a profile, without being a pretty large outlier. Kendrick has a .341 BABIP over his nearly 5,000 PA career, and it's hard to think of a more unique hitter than Ichiro.

But that's not all. This year also featured a career low for Holliday in FB%, a career high in GB/FB%, and the second-lowest Hard Hit rate. One thing to be optimistic about: he did post a career high LD%. But Line Drive percentage is actually one of the least predictable stats we're talking about here. And we are just talking about 277 PA. But these aren't minor changes here: Matt Holliday looked like a completely different hitter than the one he has been his entire career. And these stats actually don't need much time to stabilize. Fangraphs says you only need 80 BIP to get a reliable number for GB% and FB% (it's 600 for LD%). That's compared to 820 balls in play for BABIP. Holliday had 183 balls in play last year so that actually means these numbers should be considered pretty reliable.

In fact the same link says you need 50 fly balls to get a good measure of HR/FB%. Previously I felt it was easy to discount the fact that Holliday also posted a career low HR/FB% due to a small sample. But being that he hit 53 fly balls in 2015, it seems more concerning than I thought. I can't find a similar number for finding a reliable Pull%, but it seems unlikely to exceed Holliday's sample size in 2015.

Of course, these thing really exist on a scale: 50 fly balls aren't perfectly predictable while 49 are not. 51 is more predictable than 50, which is little more predictable than 49 and so on. And these numbers don't mean perfectly predictable anyways, but reliable. My point is simply that these sample sizes are actually big enough to be worth giving serious weight to.

I know what you're probably thinking right now. Holliday was hurt for parts of 2015, these numbers are tainted a little bit from that. Well, unfortunately it wasn't looking good even before his initial injury:

Holliday had 218 PA before getting injured. Over that time frame, he still was running the lowest Pull% of his career. And he actually pulled the FB% and Hard% up after coming back from injury. But, Holliday has always been a better hitter in the second half than the first. Most hitters are, because hitters generally do hit a little better when its warm, but Holliday over his long career holds a 133 wRC+ in the first half compared to 145 in the second half. Let's look at these batted ball stats over his career, separated by month.

Other than generally hitting more grounders in April and May,  his profile doesn't change all to drastically. Sure, some of the lack of fly balls can be attributed to this, but not all of it or even most of it. His 1.56 and 1.39 GB/FB for his career in April and May is dwarfed by a first half GB/FB of 1.84 in 2015. And the monthly splits don't show any discernible difference in Pull%. So this can't be blamed on First Half Matt.

I wanted to see if Holliday had been pitched different in 2015, so I headed over to brooksbaseball.net. Here is how he was pitched from 2012 to 2014, and here is how he was pitched in 2015. Taking the inside third of the plate, Holliday saw 11.9% of pitches inside and in the zone in 2015, compared to 12.2% from 2012-2014. As far as inside and off the plate, those numbers were 16% in 2015 and 17.6% from 2012-2014.

So Holliday was pitched a little less inside in 2015, which might mean that he should be going the other way a litt more often, but it certainly doesn't seem to explain a career low Pull%. Taking the outside third of the plate, Holliday saw 10.7% of pitches in that region compared to 11.5% in 2012-2014. For off the plate, it's 11% and 11.3%. In terms of where to pitch Holliday, there doesn't seem to be any change among pitchers.

There did seem to be a slight change in how Matt was pitched. His 39.5% Zone% was lower than the last two years (42.5% and 42.0%). Taking it by pitch, 61.3% of pitches he saw were fastballs last year, compared to 64% in 2014. But again, these are very small changes. These stats have fluctuated to similar degrees throughout his career without causing drastic changes in Holliday's profile.

So, this post stops short of explaining why Holliday changed so drastically in 2015. But you can't really argue with the fact that these changes happened, and over a reliable sample. And these changes are concerning ones. The Cardinals want Holliday to hit the ball with authority, and for the first time in his career there seemed to be a lot of reason to think that he wasn't. Color me concerned about the Elder Matt in 2016.