In 2016, I wrote an article about optimizing the Mariners’ lineup a couple months into the season. To make it easier (so as not to force you to open another article) here’s a paraphrased version of what I said lineup optimization means:
In economics, there is something called the rational choice theory. What the rational choice theory assumes is than an individual has preferences among choices. These preferences are assumed to be made by surveying all possible options, taking in all information, and to make the most logical decision that is of greatest benefit to them.
To put out the best lineup possible, one should use all available information. In The Book: Playing the Percentages in Baseball, Tom Tango (as well as Mitchel Lichtman and Andrew Dolphin) attempt to quantify this using empirical data. In the book, there is an entire chapter devoted to sabermetric lineup construction.
Following is a graphic with the conventional, old-school method of building a lineup, as well as an optimized, sabermetric version of how to construct a lineup. The standard version is more philosophy and trait-driven while the sabermetric version revolves around statistics (duh).
Admittedly, over the course of a baseball season using an optimized lineup only adds a couple wins. That’s relatively insignificant, and one could argue it’s trivial. But if you’re the Mariners who look like favorites to finish as the second Wild Card team, it can be the difference between playoffs and, well, not playoffs. With an injury to Drew Smyly before the season has started, little things like lineup optimization become increasingly important.
This lineup will likely look very different from the opening day and regular lineup put out by Scott Servais, and that’s okay!
Because year-to-year numbers are so different due to injuries, sequencing, mechanics, and so on, we will focus on players’ career numbers with 2016 performance and 2017 projections as context. On we go!
I came up with the following orders of lineups versus left-handed and right-handed pitchers:
(“Stat type” denotes if I used career numbers or my personal projections.)
You will notice that the optimized lineup doesn’t follow the sabermetric setup to a tee. Why? As an example, managers like to have speedy players who can get on base some at the bottom of the lineup, because it’s as if having two leadoff hitters. Also to better match certain philosophies and because some players will do better or worse than their current career numbers or 2017 projections. Namely, players like Haniger, Segura, Zunino, and Martin. In Pt. 2 of this article, I will explain why I expect players to deviate from these numbers.
Because of injuries, fatigue, “hot/cold streaks”, and maintenance days, the lineup isn’t always going to look like this. And really the lineup will probably never ever look like this. No managers in the MLB are nearly as progressive as many sabermetricians, and so even though some unconventional tactics are utilized, many still use concrete roles in the bullpen and outdated methods of lineup construction. As an example, managers are unlikely to move star players such as Seager or Cano down in the lineup even if they are significantly worse against a certain handedness of pitchers. Due to the egos and psyches of players, this may be the correct way to go about things at times, but it doesn’t always make sense. That said, all things relatively equal, it’s generally safe to go with the more experienced (or emotionally demanding) player. So Seager will unlikely bat as low as sixth like he is in my “vs. LHP” optimized lineup.
For the first time since October, there will be regular season Mariners baseball tomorrow. The Mariners play the AL West favorite Astros, and Felix will look to recover from his past two seasons in which he has failed to look like King Felix. Last season he really failed to look like Felix Hernandez. I can support the blonde beard. Not so much the slicked back blonde hair.
Pt. 2 of my article explains my lineup construction briefly and the Mariners season in context!