If you’re unfamiliar with the concept of lineup optimization, you’re not alone. From the time we’re little it’s drilled into our heads that baseball lineups are already at their most successful state when using the following logic, and the one of the lineups proposed:
1. Speed guy (Segura, SS)
2. Slap hitter (Lucroy, C)
3. Best overall hitter (Braun, RF)
4. Best power hitter (Ramirez, 3B)
5. 3rd best hitter (Gomez, CF)
6. Best remaining power hitter (Reynolds, 1B)
7. Best remaining hitter (Davis, LF)
8. Last position player (Gennett, 2B)
If you look around baseball, that’s generally what you’ll see. However, what if I told you that lineups were not being optimized by Major League managers? Well, it’s pretty much true. Despite the integration of advanced analysis in front offices, the sentiment hasn’t fully flourished in the dugout. But then, how would a lineup be optimized, and then what would the Brewers lineup look like in that state? I used a lineup optimization simulator using Dan Syzmborski’s ZiPS projections as my offensive output numbers.
To begin, one must understand the general theory behind an optimized lineup. Note that this is not merely philosophy, but backed up by data and the running of a great number of simulations. An optimized lineup generally looks like this:
1. Best OBP
2. Best overall hitter
3. Best remaining hitter after 1, 2, 4, and 5
4. Best power bat
5. Best remaining hitter after 1, 2, and 4
6. Best remaining hitter after 1, 2, 3, 4, and 5
7. Next worse hitter
9. A low power hitter, who is not good, but can get on base enough.
So how does the Brewers’ lineup look with this arrangement? I’ll admit that I was surprised when I saw the result. This lineup would be projected to average 4.72 runs per game which would’ve placed them fifth in all of baseball last year, and an improvement over the 4.525 per game that the traditional lineup at the beginning of the post is projected to produce.
1. Jonathan Lucroy, C
2. Ryan Braun, RF
3. Jean Segura, SS
4. Mark Reynolds, 1B
5. Aramis Ramirez, 3B
6. Carlos Gomez, CF
7. Khris Davis, LF
9. Scooter Gennett, 2B
The most shocking result was Jonathan Lucroy leading off. However, of all the top optimized lineups, Lucroy was the leadoff hitter in most of them. Ryan Braun hitting second should be no surprise to anyone familiar with the sabermetric optimization. Jean Segura at third also makes sense. I was a bit surprised with Mark Reynolds hitting fourth over Aramis Ramirez. This may be in part due to ZiPs being relatively high on Reynolds compared to most people’s projections, but also that Reynolds is younger, and less likely to see severe skill deterioration. The rest of the lineup is not all that shocking unless you’re a strong opponent to the pitcher hitting eighth (which only is projected to gain two additional runs per year, on average).
So what does all this mean? Pretty much nothing. Ron Roenicke will never employ this lineup over the traditional one nearly every manager uses. The potential backlash from the media, most fans, and his own players could make it not worth the change. However, in a vacuum, it would help the Brewers score more runs. The optimized lineup would score about 31.59 runs more in a year. With the general rule of ten runs being equal to one win, and optimized lineup would get the Brewers about three more wins this year than the traditional one.
Tags: Milwaukee Brewers