Optimizing The 2023 Red Sox Lineup

Every season, there are frequent fan arguments over what would be an optimal lineup for their preferred team. Tools on Baseball Savant and other websites or programs provide some insight into how to build a lineup by providing a breakdown of a player’s performance in each specific lineup spot, but this still leaves a lot up to the unknown. Since 2019, the Red Sox have had rather abysmal luck finding a consistent, fit leadoff hitter.

In the first half of 2022, the Red Sox mainly called upon Enrique Hérnandez with sprinkles of Jarren Duran before pivoting to Tommy Pham after the Aug. 1 trade deadline: placing him in the leadoff spot for 45 of his 53 appearances with Boston.

Hérnandez slashed .218/.275/.376 from the one-hole, as Duran posted a .239/.254/.362 line of his own. Later on, Pham slashed .239/.305/.375— which at least qualifies him as the only primary Red Sox leadoff hitter with an OBP in the 300s!

Additionally, Duran had a 27% strikeout rate in this spot: the leadoff spot is where a strikeout causes the most damage throughout the entire lineup. Pham had a K rate of 28%.*

*A mark of 20% is considered average, anything over 27% is, in the kind words of Fangraphs, “Awful”

The woes of these leadoff hitters spurred my further research into this topic.

Although there are plenty of stereotypes in baseball surrounding the idea of building a lineup, sabermetrics and recent advances in lineup optimization research over the past decade have initiated a new area of development— setting a new standard of excellence throughout the major leagues. One of these non-digital advances is Tom Tango’s The Book.

Originally published in 2006, The Book includes a hefty chapter on lineup optimization that looks at factors such as DP%, baserunning, wOBA, and other statistics that can be used to build an optimal lineup. The majority of my research has come from this chapter of The Book, and will be included to find the optimal lineup for the 2023 Red Sox.

Baserunning, Base Stealing, And Their Paradoxical Tendencies

First and foremost, the task of optimizing a lineup cannot commence without deeper research into what factors result in optimal performance for the players in each individual lineup spot. Of these factors, run values by event and value (modified by plate appearance (PA)), strikeouts, and double-play tendencies— that is, how often specific players hit into double plays— provide the most significant insights on how to construct an optimal lineup; and how to squeeze out as many extra runs, and wins, for a team as possible. Notably absent from these important factors are baserunning and base stealing — this is because building a lineup to favor individual base stealing basically cancels out the benefits of skilled baserunning. To better explain this, here is a quote from The Book itself:

“A great baserunner is usually, but not always, a great basestealer. …[Y]ou would prefer to steal a base, and get into scoring position, as long as you have a hitter that hits a lot of singles, but not many extra-base hits. This is the reason we want our basestealer to bat fifth or sixth. At the same time, you prefer to have your great baserunner in front of hitters that hit a lot of singles [in order to optimize their value], regardless of their power numbers. This is the reason we want them at the top of the order.

Essentially, this creates a paradox where we are trying to put the same person— one that is a good baserunner and base stealer— into opposite positions in a lineup, which is why worrying about speed in its different aspects is a bit of an insignificant factor to consider when optimizing a lineup (this is especially true for our purposes, as the Red Sox themselves don’t boast fantastic baserunners or base stealers).

Plate Appearance Factors & How They Affect Run Expectancies

The most important thing needed to understand the factors that contribute to the practice of lineup optimization is a table of run values by event and value as modified by PA. But before I show you this single—terrifying— chart, it is important to understand that a lineup is a continuous loop; statistically, it has been evaluated that each lineup slot gets about 2.5% more plate appearances per game than the next (as shown by the table below). These factors will affect run value per slot, by either having a greater impact on the overall game— as seen in higher batting slots— or taking significance away from individual plate appearances— as is the case for lower batting positions.

Table from Tom Tango’s The Book

*BOP= Batting Order Position

The concept of a lineup being a continuous loop is initially important to understand: as the repetition of a lineup—rather than it being a single, static, one-time event— has an effect on run values for each event in each lineup spot. In effect, this table basically means that each previous lineup position has 2.5% more of an effect on the outcome of any game than the following spot. In this chart, the five-spot is set as an arbitrary zero in this chart, meaning that events in the four-spot are 2.5% more impactful than if the event took place in the five-spot, while those in the six-spot are 2.5% less impactful, and so on.

Understanding this table and understanding the effects that lineup position has on an event’s impact is necessary to understand the big picture of these combined contexts. It is important to factor in PA value, along with run expectancy (RE) when building a lineup so as to fully understand the value of each event in each position.

Now if you thought that last chart was something, please hold your applause, as I am about to show you the monstrosity of a chart that I mentioned before that combines these two aforementioned concepts into an end-all-be-all lineup optimization chart.

Table from Tom Tango’s The Book

*Note: This table is based on real-life American League numbers, not simulated projections

I told you it was terrifying, didn’t I?

Here are some general distinctions, mentioned throughout the lineup optimization chapter of The Book, that should be taken from this monstrosity of a chart:

No. 2 and No. 4 Hitters

The No. 4 hitter has an advantage in all XBH categories, but the No. 2 hitter has an equal advantage in NIBB and HBP. This means that the No. 4 and No. 2 hitters should be relatively equal in performance, with the No. 4 player boasting a better SLG%, while the No. 2 hitter should provide better OBP numbers.

Another thing to keep in mind with this is that the No. 2 hitter will have 5% more plate appearances than the four-spot batter, but on average the No. 4 hitter will be up with more men on base.

No. 2 No. 3, and No. 4 Hitters

Additionally, as the No. 2 and No. 4 hitters are relatively equal, they also have similar run value advantages over the No. 3 hitter (~0.02-0.03 runs). This means that the No. 3 hitter should be worse than both the No. 2 and No. 4 hitters.

These adjustments, when correctly implemented, can yield an additional ~0.02 runs per plate appearance at most. While this probably seems unimportant and easy to overlook, over the course of a year, this advantage adds up to approximately 10-15 runs, which can result in up to four more team wins due to proper optimization.

Strikeouts And Their Effects On Optimization

It is also important to organize the lineup in order to maximize strikeout effectiveness. What this means is, when looking at the RE by event and value/PA table, you want to place your hitters with the highest strikeout percentage in spots where striking out is less detrimental. Similarly, it is also ideal to place your better-disciplined players in spots where the strikeout is more costly.

In this example, the three, four, and five spots are on opposite ends of the spectrum. Because of this, you want your best hitter with the most strikeouts in the three-spot, so as to maximize run value. This is because when the No. 3 hitter strikes out, it has less effect on the game than if the No. 1-5 spots were to do the same.

Additionally, it is best to put a low strikeout hitter in the fourth and fifth slots, where strikeouts have some of their most costly values.

This is a secondary reason why it’s best to have Devers slotted third instead of fourth, as this spot in the lineup has the smallest negative strikeout impact of any of the first four spots in the order.

Double Play Tendencies Throughout A Lineup

Double play (DP) situations per game, relative to league average by batting order, are important to consider when building a lineup: especially in the No. 3 (0.18 DP situations/game) and No. 5 (0.02 DP situations/game) spots — where the numbers of DP/game are at their separate extremities. For this quantification within the Red Sox lineup, I evaluated DPrate, which rates a player’s tendency to hit into double plays relative to league average (2022’s league average for DPrate was 10.1%). The player with the best DPrate was placed third, as they are less likely to GIDP, while the best player with the worst DPrate was placed fifth, where they experience fewer chances to hit into a double play. I calculated this value for each main Red Sox hitter in 2023 (and Trevor Story).

Creating An Optimal Red Sox Lineup

With all of this research, I put the information in context with this year’s Red Sox roster to create the optimal lineup. I will insert the chart below and then explain the position of each player following the lineup.

Leadoff: Masataka Yoshida

Although the information for Yoshida is limited compared to other players due to less available data, he is genuinely the best leadoff candidate that the Red Sox have employed in a long time. Yoshida boasted an OBP of .447 in 119 games. The current Red Sox player with the highest OBP last year is Rafael Devers, who ended 141 game season with a .358 mark.

Yes, you are reading that correctly, Yoshida posted an OBP almost 100 points higher than the next closest Red Sox. Although this should be taken with a grain of salt since Yoshida was playing in the Japanese Pacific League, his plate discipline should transfer over relatively well to MLB play.

So, why does this make him such a good candidate to lead off? In the one-spot, walks and singles are of equal value, meaning that anyone with a high OBP is a good idea. Additionally, he had a strikeout percentage of 8.1% last year and a mark of 9.4% for his career. This is another factor that places him in this spot, as strikeouts have the most negative impact on the game within the first three hitters of a lineup.

No. 2: Justin Turner

Your best hitters should be placed somewhere among the one-, two-, and four-spots. It is counterintuitive for me to say this but after Rafael Devers… Turner looks to be the most well-rounded Red Sox hitter. Turner is in the two-spot instead of the four-spot because of this, and because of his high OBP and ability to drive in runs (SLG%, wOBA).

As previously mentioned, the No. 2 and No. 4 hitters should be relatively equal with the No. 4 hitter providing better power numbers, which is the perfect description of Justin Turner and Rafael Devers.

No. 3: Alex Verdugo

One thing almost kept me from putting Verdugo in the three-spot, and that was the fact that his strikeout percentage is the lowest among returning players, and the No. 3 spot has the least detrimental strikeout value among the top half of the lineup. But, all things considered, the hitters I have placed fourth and fifth still have above-average percentages, so Verdugo was still the best fit here.

Additionally, as previously mentioned, the three-spot sees the most double play situations per game on average, and Verdugo’s stat at 9.9% (above league average), which is especially great in comparison to other three-spot candidates such as Enrique Hérnandez.

Finally, Verdugo ranked fourth among returning players last year in SLG%, and the home run has the second-highest value in the lineup in this spot.

Cleanup: Rafael Devers

Devers’ above-average strikeout plays well in this high-leverage spot, and his power and ability to drive in men on base don’t hurt either. Having Devers as a cleanup hitter should be preferred, as the spot sees the highest advantages in the values of all extra-base hits. Devers led all Red Sox in 2022 with 289 total bases, so this advantage would be best utilized by having his experience in this spot.

No. 5: Triston Casas

The main reason why I decided on Casas for this spot is due to what I mentioned previously about double-play factors within a lineup. Although he did not have a whole season to prove himself last year, Casas had the highest double-play rate among Red Sox at 23.5 (the next closest were Hérnandez and Adalberto Mondesi at 16.7%), grounding into four double plays in the 17 double-play situations that he came up to bat in— in 27 games played in the majors.

Another concern would be Casas’ below-average strikeout percentage since the No. 5 spot does have a high strikeout disadvantage.

That being said, I am a Casas believer and definitely see him as being a large contributor for Boston this season. He has a more complete hitting profile than the rest of the team, even if it is in the minor leagues, and doesn’t deserve to be stuck lower than No. 5.

No. 6: Adam Duvall/Bobby Dalbec

Although last season was definitely not the best for Duvall, I do have relative faith for him to return to being a replacement-level player, which is all anyone can really request from a No. 6 hitter.

Dalbec carries the same sentiments for me. I am a big proponent and cheerleader for him, but do not expect top-half-of-the-lineup production from him after seeing his 2022 numbers.

No. 7: Christian Arroyo/Rob Refsnyder

Arroyo actually had a statistically better season than Adam Duvall last year!

Though that may be a surprise to most — it was to me — I still put him below Duvall since Arroyo has less of a threat for power. That being said, Arroyo and Duvall did boast similar numbers last season (with Arroyo having a better SLG% over the year than Duvall), and if he can keep this up in 2023 I would definitely move him to the six-spot.

As for Refsnyder, he was also quietly impressive last year, but with an obvious lack of power threat. He should give good consistency to the bottom of the lineup and I hope to see him while the Sox work through some injuries.

No. 8: Enrique Hérnandez/Adalberto Mondesi

Hérnandez has consistently been placed at the top of a lineup, where he really doesn’t belong. Hopefully, that trend stops this year, as his OBP, high K%, and below average DPrate have proven costly in higher spots.

As for Mondesi, I have really no clue what to expect of him. But, I do expect him to be better than whatever catcher Boston decides to roll with.

No. 9: Reese McGuire/Connor Wong

McGuire is a step up from the abysmal nine-hole hitters that the Red Sox have had in past years (ahem, Franchy Cordero). Still, though, he should not be placed any higher in the lineup.

It’s nice for the Red Sox to have someone with an over .300 OBP at the bottom of the lineup, though!

Qualifying The Analysis

Of course, the implementation of this optimal lineup is up to Red Sox officials, staff, and managers. That being said though, the Sox lineup construction these past few years has angered me like no other, so hopefully, they take at least some of these optimization factors into consideration for the 2023 season.

I would say that now, more than ever, these ideals are pertinent to implement. Boston arguably got worse over the off-season, but they could do better than they are currently predicted to do with an optimized lineup. As previously stated, lineup optimization can result in four or five more wins for a Red Sox team that is slated to finish last in a competitive AL East.