Diamondbacks’ Platoon Advantage

What are platoons? Two players share a position. There are several types. Three examples are:

• old & declining player paired with a young inexperienced prospect.
• strong batter & poor fielder paired with a weak batter & excellent fielder.
• left handed batter paired with right handed batter.

This article focuses on the third example.

What is platoon advantage? A left handed batter hits best against right handed pitchers (opposite handed pitchers), while a right handed batter hits best against left-handed pitchers. Two possible reasons that batters hit best against opposite handed pitchers:

• Batters have a better view of the approaching baseball.
• Breaking balls tend to break away from same-handed batters (and are harder to hit), while they tend to break toward opposite-handed batters (and are easier to hit).

Platoon advantage was confirmed in a study. It assumed pitchers with equal skill. Further, this difference was larger for marginal batters than for the best batters. For details here is the study.

How would the platoon work? When the starting pitcher is right-handed the left-handed batter would start the game, while the right-handed batter would be on the bench. When the starting pitcher is left-handed, the roles of the platoon players are reversed. The impact is increased with more at-bats with the batter opposite-handed from the other team’s pitcher.

In the last two seasons, 72% of Diamondback plate appearances were against right-handed pitchers and 28% were against left-handed pitchers. This proportion is not unusual – in the Majors 29% of starts were by left-handed pitchers and 28% of the innings pitched were by left-handed pitchers. Because most pitchers are right-handed, the platoon’s impact is highest when the left-handed player in the platoon gets the most plate appearances.

“Platooning has largely become a casualty of the expanded bullpen that came into vogue in the 1990s.” – Steve Treder.

Looking ahead, platoons may increase. Previously, it was easier to counter platoon effectiveness by bringing in a same-handed pitcher from the bullpen, even for just one batter. The 3-batter minimum for bullpen pitchers greatly reduced this counter-strategy against the platoon.

If the roster size increases above 26 players when a new Collective Bargaining Agreement (CBA) is negotiated between the MLBPA and MLB, that extra player (or players) could become part of a platoon.

Did the platoon advantage apply to the Diamondbacks in 2019 and 2020?

For 2019 and 2020, let’s look at three measures of batter results (HR/PA, OPS, and wRC+). For each year, the results will be presented to show differences between a) left-handed at-bats vs right-handed at-bats, and b) platoon advantage vs no-platoon advantage. Results shaded in green mean platoon advantage was positive. Results shaded in red mean platoon advantage was negative.

Diamondback at-bats (excluded at-bats as non-Diamondbacks). Data from FanGraphs.

The table shows that in 2019 the platoon advantage results were favorable for HR/PA, OPS, and wRC+. Building on that success, the next season the Diamondbacks wisely increased the percentage of plate appearances with platoon advantage from 61.4% to 65.3%.

However, an unexpected surprise happened in 2020. For right handed at-bats vs left-handed pitchers, the results (HR/PA, OPS, and wRC+) were not better. HR/PA was about the same, while OPS and wRC+ were worse. This anomaly called for a closer look.

“[Compared to pitchers, batters] ...are much, much quirkier, and they put their own stamp on things.” —Graham MacAree.

A closer look at right handed at-bats in 2020. Let’s look at batter statistics to understand possible reasons that right-handed at-bats did not realize a platoon advantage in 2020.

2020. Excluded non-Diamondback at-bats. Data from FanGraphs.

The last three columns of following table would be much more green if the platoon advantage had occurred.

2020. Excluded non-Diamondback at-bats. Data from FanGraphs.

Possible reasons that a platoon advantage was not realized for right-handed at-bats:

Batters were unfamiliar with the specific left-handed pitchers. The unfamiliarity reduced performance by an amount greater than the platoon advantage. Supporting this reason are two points:

• Five of the eleven batters had less than 300 PAs in the Majors at the start of the season. Two of the five made their Majors debut this season. Therefore it is likely that they were unfamiliar with the left-handed pitchers they faced in the Majors.
• This article quantifies the unfamiliarity effect of left-handed pitchers as greater than the platoon effect. It’s a very long article, so you may want to skip to the next to last last table, which is the quick summary.

Andy Young and Tim Locastro hit extraordinarily well against right-handed pitchers (no platoon advantage). I tip my hat to the Diamondback batting coach. Whatever they did, their positive results had a larger magnitude than the platoon advantage. Those results were over 38 PAs, and may not be sustainable over 162 games.

The platoon effect was weaker for Diamondback switch hitters. Although two of the three switch hitters demonstrated a platoon effect, their HR/PA rates were low. Another way to say it: Switch hitters were in a batting slump. Excluding right handed at-bats by switch hitters would have turned HR/PA into a platoon advantage. The following shows the batting slump compared to last season:

• Ketel Marte. For all plate appearances, his OPS fell from .981 to 732, and his wRC+ fell from 150 to 95.
• Eduardo Escobar. For all plate appearances, his OPS fell from .831 to .605, and his wRC+ fell from 109 to 56.
• Ildemaro Vargas. For all plate appearances, his OPS fell from .712 to .536, and his wRC+ fell from 82 to 39.

The platoon effect was less than random variation. It was a short season; with few PAs statistics can show wide fluctuations/variations (source article). This article has a calculator to determine the number of PAs needed to have 50% confidence in wOBA. It shows that 400 PAs are needed. For the season, all the right-handed batters had less than 400 PAs. Their season statistics may have fluctuated wildly because the small number of PAs.