Two terms I’ve heard thrown around so frequently between the front-office, the coaching staff and throughout the Twitter-sphere that defines the Cardinal’s lineups year-after-year are ‘flexibility’ and ‘consistency’. Matt Carpenter’s career alone is anecdotal of this see-saw dynamic. When Carpenter began playing regularly in the majors in 2012, he originally served as a utility-type player seeing time at first, second, third, left field and right field (flexible). Then 2013 saw Carpenter transition to the every-day second-basemen logging 74% of his games there. 2014 saw Freese traded away and Carpenter shifting to third-base now logging 95% of his games at third for the 2014 and 2015 seasons combined, back to consistency. Then 2016, the Cardinals acquired Gyorko and Carpenter saw 39% of his games at third, 32% at first and 29% at second. Last year, again, shifting from flexibility to consistency Carpenter saw 79% of his games at first, and yet this year Carpenter has split his time between third and second. It seems every-time Mozeliak and Matheny try to embody one philosophy, they pivot back to the other when things start getting muddy. My goal with this exploration isn’t to play judge or jury of which side of the dichotomy is the ‘right’ one, but rather to explore Matheny’s strategy with the Cardinals lineups since his debut in 2012… with that, welcome to Matheny Under a Microscope, Chapter 2: The Lineup Shuffle.
The first model I wanted to analyze was the general trend of whether teams with more or less lineup flexibility had any correlation with a winning season (using data from the 2017 season). After running a bunch of different permutations of total unique batting orders, total unique defensive lineups, wins, runs/game, players used, and total bases, there truly wasn’t a model that justified a ‘correct’ strategy. In 2017, the league average for unique batting orders was 121 and the league average for unique defensive lineups was 99; the Cardinals we’re above league average in both categories at 144 and 116, respectively. One observation to be made was that 9 of the 10 playoff teams from 2017 were above average in both categories as well, the Rockies being the exception. This leads me to believe that some degree of roster flexibility and depth may translate to more wins, especially when considering how long the season is along with the potential for injuries.
From here out I’m going to be focusing on offensive batting orders, regardless of defensive positioning.
Matheny has had an interesting trend of total unique batting orders used represented in the following table:
Matheny started his tenure with 122 unique batting orders, dropped down to 89 his second season and has seen that number rise or plateau each year. Again, there’s no real trend in recent years between total unique batting orders and wins or runs, but an interesting comparison is Tony La Russa. During La Russa’s 16 seasons he averaged 131 unique batting orders and 88 wins, pretty similar to Matheny's averages (126 and 91).
At this point, I wanted to explore some trends in Matheny’s batting orders in hopes to understand his thoughts and whether or not he may actually be ‘optimizing’ it, a term so often coined by none-other than John Mozeliak.
The two metrics I’m going to use to analyze Matheny’s batting orders are wOBA and wRC+. I’m not going to attempt to simplify an explanation for these metrics, thus if you are unfamiliar I suggest reading up on them from their respective websites on Fangraphs to get the best idea of how they are calculated. Of note, wRC+ is standardized around the value 100, thus a value of 110 is 10% better.
With the formalities out of the way, let’s jump-in. First, to set a nice benchmark on what a long-term 'traditional lineup' might look like, here is the metrics for Tony La Russa’s lineups from 2002-2011 (Unfortunately Fangraphs doesn’t have the splits dating back to the beginning of his tenure).
Some observations to be made here: the strongest part of the lineup revolved around the third spot in the order and gradually decreased in both directions. I know what you might be thinking, yes Albert Pujols (I’m sure you’ve heard of him), spent 84% of the time hitting in that 3rd spot during this time and explains the borderline ‘elite’ level of production. But this exercise isn’t to indulge La Russa, simply to use him as a benchmark.
If we analyze Matheny’s career (2012 – 2017), his metrics measure out relatively parallel to La Russa’s:
Of note here, Matheny has never had a batter with the same talent-level as Albert Pujols and the production is a step-down especially in the 3rd and 4th spots, but there is somewhat parallel trend: Matheny over his career has managed to allocate the most at-bats to his three best hitters, by generally placing them early in the lineup. Where Matheny’s judgment begins to come into questions is of late, specifically in 2016 and 2017. I’ll start with 2016:
The 2016 lineup was borderline anemic in the 3rd and 4th spots, most frequently manned by Matt Holliday and Stephen Piscotty, respectively. An admirable attempt to give Holliday a farewell-tour season and Piscotty a chance to thrive, but the two ‘power’ hitters combined for 42 home runs. Meanwhile, the most home runs that season were produced by Jedd Gyorko, who in a limited role clobbered 30, while mostly hitting towards the back end of the lineup, evident by the increased production in those spots. Now, it’s hard to blame Matheny as Brandon Moss and Randal Grichuk had power potential, with a combined 52 home runs that year, but also a combined 282 strikeouts, not ideal for the 3rd or 4th spot. Jedd Gyorko perhaps could’ve been serviceable higher up in the order, but 2017 might tell us otherwise.
2017 presents an even-more deficient lineup with really two standouts, the 1st and 2nd spots in the order, mostly manned by Carpenter and Pham respectively, to nobody’s surprise. What is notable is the 5th spot in the order is the least productive. Yes, there is an argument to be made that the batter hitting 8th, regardless of who, benefitted from hitting in front of the pitcher, but this lack of production from the 5th spot is a result of Matheny’s biggest deficit: looking backward and not forward. Yadier Molina, yes beloved Yadi, occupied the 5th spot more than anyone else, spending 70% of the season there. We all love Yadi, and if you don’t think you do, you’re just lying to yourself, but Yadi is not the man to hit in the 5th spot, not last year at least. Yadi had a modest .273/.312/.439 season, nothing to be disappointed about for a catcher in his age 34 season with his intangibles, but we can’t overlook the fact that his average and on-base percentage were down from both the year before and his career numbers. Who would’ve been better suited? Perhaps Gyorko, DeJong or J. Martínez but rearranging the lineup just has a cascading effect that at the end just isolates the lack of a truly feared hitter: one that Ozuna has hopefully filled.
Thus far in 2018, Matheny has essentially rolled out the same lineup daily with Fowler, Pham, Carpenter, Ozuna, J. Martínez, Molina and DeJong in that order followed by a combination of Gyorko, Wong and Munoz in the 8th spot and the pitcher batting 9th. Now there is an argument to be made about DeJong hitting 6th and Yadi down to 7th, but the season is early, and we’ll have to wait a bit longer before making any type of judgement. DeJong started off as arguably the hottest hitter on the team, but then proceeded to strike-out in 6 straight at-bats over the course of 2 games. There are certainly permutations of the top 5 guys as well, whether that be flipping Carpenter or Fowler, or perhaps rolling out Carpenter, Pham, Ozuna, J. Martínez, Fowler in order to get Ozuna more at-bats, but Fowler deserves some time to establish himself at the top this season in my book.
To recap, Matheny has generally seen an increase in total batting order permutations throughout most of his tenure, but of late he has been placing declining-franchise players in premiere positions because of their past and not so-much their current ability to produce. I do believe there was some lineup optimization that was not taken advantage of, especially the last two years, but now that Matheny’s finally has a Pujols-esque presence in Ozuna, he will be truly tested to optimize the lineup this season.
*Thanks as always to Fangraphs, Baseball Reference and @cardinalsgifs for the goods.