Assembling the Ideal Cardinals Lineup

Few aspects of the game polarize conventional and sabermetric baseball rhetoric quite like the daily lineup card. It shouldn't come as any sort of surprise that I fall into the latter school of thought. Everything about lineup construction embodies the sabermetric revolution that has dominated the 21st century: it's nerds like myself armed with spreadsheets and data tables looking to take the provided resources and squeeze out as many runs–and subsequently wins–as possible. If we abide by the rule of thumb that roughly ten runs equates to one win, even the most aggressive estimates would place the value of lineup constructions at a couple extra wins a year. And while a whole article dedicated to the arrangement of the starting nine may seem niche, just ask the 2016 Cardinals if the ability to snag a competitive advantage, however slight, makes a difference.

The methodology

Regarding the actual thought process behind assembling the lineup, I deferred to The Book: Playing the Percentages in Baseball, in which Tom Tango, Mitchel Lichtman, and Andrew Dolphin devote an entire chapter to lineup optimization.

However, I still needed a way to quantifiably evaluate and compare the Cardinals batters. To create a more balanced assessment, I mixed together the ZiPS and Steamer rest-of-season projections–collectively known as Depth Charts projections–along with players' year-to-date stats. To combat year-to-date numbers tainted by small sample sizes, I took the number of plate appearances a player accumulated over the past calendar year and subtracted that from 1500. (I chose 1500 so players with a sufficient sample of about one full season would have their year-to-date stats, ZiPS projection, and Steamer projection weighted roughly equally.)

Take Dexter Fowler. Fowler has tallied 549 plate appearances over the past year with a wOBA of .353, leaving the remaining 951 plate appearances using his projected .342 wOBA. Average those two together and we get a final aggregated wOBA of .346 for Fowler. I repeated this process using various statistics for each position player currently on the active roster.

Choosing the starters

To determine which eight position players would make the lineup, I ranked each player by fWAR/600 plate appearances using the same method described above. That gave us the following group:

Catcher: Yadier Molina

First Base: Matt Carpenter

Second Base: Kolten Wong

Third Base: Jedd Gyorko

Shortstop: Paul DeJong

Outfield (arrange these three however you please): Dexter Fowler, Tommy Pham, Randal Grichuk

The lineup

(Note: All "counting stats" below are on a per 600 plate appearances scale.)

To construct the optimal sabemetric lineup, it is recommended that you begin by singling out your top three batters. Based on the wOBA column, that trio consists of Carpenter, Pham, and Fowler. The player with the highest OBP should bat leadoff, making Carpenter the easy choice here. The other two should hit second and fourth, with more walks in the #2 hole and more extra-base hits in the cleanup spot. Knowing this, we can conclude that Fowler fits the #2 profile better than Pham, who is better fit batting fourth. So far, everything has been fairly straightforward.

Your next two most productive hitters should be slotted third and fifth. Generally speaking, the better member of that pair should bat fifth (consider how many times the #3 hitter bats with two outs and nobody on). That said, I am going to act in defiance of the wOBA column and advocate for Gyorko to hit third and Wong fifth, against righties that is. Home runs are more valuable and strikeouts are less costly in the third spot while a higher OBP is more important in the fifth spot. The one downside to this decision is that the #3 slot encounters more double play opportunities than the #5, and, as you would expect, Gyorko is more prone to grounding into a double play than Wong. Against lefties I would bat Gyorko fifth and DeJong third, moving Wong down in the order to accommodate for his disadvantage in lefty-lefty matchups.

To fill out the remainder of the lineup, place your best remaining hitter sixth, your second best remaining hitter seventh, and so on–with one notable exception. I am referring to the hipster batting your pitcher eighth trend, otherwise known as the "second leadoff hitter theory". Various simulation models have proven that this strategy of batting an actual position player directly ahead of the top of the order tacks on a few extra runs per season for NL teams, and, hey, more runs equals more goodness. Think of the lineup as a circular loop, where having more runners on base when your best hitters come to the plate is always beneficial.

So what does the bottom of the order look like for St. Louis? By this logic, 6-7-8-9 should go DeJong-Grichuk-Pitcher-Molina versus righties and Wong-Grichuk-Pitcher-Molina against southpaws.

Just to recap:

Again, will following this formula turn an 80 win team into a 100 win behemoth? No. But could it be the difference between sitting on the couch in October and slipping into the postseason with a shot at winning it all? It just might be.