Baseball Platoon Split Calculator

A .961 OPS against lefties and .692 against righties is a .269 split

In baseball, a split is one hitter's record divided by some condition, and the most common one is the hand of the pitcher: left-handed (LHP) or right-handed (RHP). The calculator above opens on a right-handed hitter with 150 AB, 46 H (10 doubles, 1 triple, 9 home runs), 20 BB, 3 HBP and 2 SF against left-handed pitchers, and 380 AB, 92 H (18 doubles, 1 triple, 12 home runs), 30 BB, 4 HBP and 3 SF against right-handed pitchers. Against lefties he hits .307/.394/.567 for an OPS of .961; against righties it is .242/.302/.389 and an OPS of .692. That is an OPS gap of .269.

All right-handed hitters in the 2026 MLB regular season posted a .741 OPS against lefties and .697 against righties, a gap of .044. This hitter's gap is about six times that, but it rests on only 175 plate appearances against lefties. Using the method from The Book, those 175 PA are blended with 2,200 PA of the average right-handed split, so only 7% of the observed gap counts, and the sample-adjusted skill gap shrinks to .061 toward left-handed pitchers. The calculator shows both numbers, and on the split plane an arrow slides the observed point to its adjusted position.

Why look at platoon splits

FanGraphs explains that most hitters do better against pitchers who throw with the opposite hand than against pitchers of their own hand, and calls this the platoon split. Unlike a day-of-the-week split, which is mostly noise, the handedness of pitcher and hitter really does shape plate-appearance results. That is why platoon splits come up in situations like these.

• Platoon lineup decisions: when a team pairs one hitter for left-handed starters with another for right-handed starters, the question is whether the gap is real skill. If the observed and adjusted gaps differ a lot, the sample is still too small.
• Late-inning matchups: when a left-handed reliever comes in, whom do you send up, and in front of which hitter do you bring in a lefty? The record against LHP is what you check.
• Fantasy daily lineups: if you set your lineup around each day's probable starter, the pitcher's hand is the first thing to check. Put candidates in the comparison list and they appear side by side as numbered points on the plane.
• Checking a broadcast stat: a line like "hitting .350 against lefties" is worth a second look. How many plate appearances is that, and is it really big next to the league's average split?

How to use it

1. In the top row, choose the batter's side: right-handed, left-handed or switch. The verdict sentences, the league-average ticks on the mirror and the regression baseline all follow that choice.
2. In the record box, fill the "vs LHP" row and the "vs RHP" row with AB, H, 2B, 3B, HR, BB, HBP and SF. On a narrow screen the box turns into a vertical mirror of [LHP | stat | RHP].
3. If you do not know the sacrifice flies, leave SF blank. It counts as 0, so OBP can come out slightly higher than the real figure. A cell also accepts a sum such as 312+64, so you can combine several seasons.
4. On the left-right mirror, the bars against left-handed pitchers extend to the left of the center axis and those against right-handed pitchers to the right. The longer side is drawn in the darker fill, and the dashed ticks are the 2026 MLB averages for hitters of the same side. The band along the top shows how the plate appearances split between the two sides.
5. To compare several hitters, type a name and press "Add to list". The list holds up to 10 players and disappears when you reload. The link made by "Copy link" carries the records and batting side currently entered.

How it is calculated

1. Singles and total bases: singles = H − 2B − 3B − HR; total bases = singles + 2 × 2B + 3 × 3B + 4 × HR.
2. AVG = H ÷ AB and SLG = total bases ÷ AB. If AB is 0, nothing is calculated and a dash is shown.
3. OBP = (H + BB + HBP) ÷ (AB + BB + HBP + SF). As in the MLB glossary, reaching on an error or a fielder's choice does not count, and sacrifice bunts are left out of the formula.
4. OPS = OBP + SLG. The two values are added before rounding, then rounded to three decimals.
5. The left-right gap is the difference between the two values you can see on screen. Plate appearances are counted as AB + BB + HBP + SF (sacrifice bunts and catcher's interference are not included).
6. The total line adds the counts from both rows and then recalculates. It is not the average of two batting averages.
7. The sample-adjusted gap: league gap + w × (observed gap − league gap), where w = PA against LHP ÷ (PA against LHP + k). The constant k is 2,200 PA for right-handed hitters and 1,000 for left-handed hitters (the regression amounts from The Book). The 600 PA used for switch hitters is an approximation borrowed from the point where their split stabilizes.

MLB average splits by batting side (2025 and 2026 regular seasons)

The records against left-handed and right-handed pitchers were pulled from the MLB Stats API player by player and added up by each hitter's registered batting side. The side totals add up exactly to the totals for all 30 teams. The gap is OPS against LHP minus OPS against RHP, and the PA share is the portion of that side's plate appearances that came against left-handed pitchers.

In both seasons the left-handed hitters' split is larger than the right-handed hitters'. Lefty bats had an OPS .081 higher against right-handers in 2026 and .096 higher in 2025, while righty bats had an OPS .044 and .021 higher against left-handers. The average wOBA splits Klaassen worked out from 2007 to 2009 point the same way: about 8.6% for left-handed hitters and 6.1% for right-handed hitters. Lefty bats also see fewer left-handed pitchers: 24.9% of their 2026 plate appearances came against LHP, versus 34.2% for righty bats. The right-handed split doubled from .021 in 2025 to .044 in 2026, but that is one season, too soon to call a trend. Switch hitters usually bat from the side opposite the pitcher, so their average gap is small, between .003 and .018.

A small-sample split gets pulled back toward the average

FanGraphs stresses that an observed split should not be treated as the player's true split. In a 300-PA split, only about 75 to 100 PA come against lefties, which tells you almost nothing about the hitter. The Book (Tango, Lichtman and Dolphin) handles this with regression. As Klaassen (2010) summarizes it, a left-handed hitter's record is blended with 1,000 PA of the league-average split against left-handed pitching, and a right-handed hitter's with 2,200 PA. Lefty bats are regressed less because their true skill differences are larger. No source found gives a regression constant of the same kind for switch hitters. Mailhot (2022), citing the authors of The Book, reports that a switch hitter's split stabilizes at about 600 PA against LHP, and the calculator borrows that point as its constant. It is this tool's approximation, not a published regression amount.

Even an everyday player with 600 PA a season would see only about 205 PA against lefties as a right-handed hitter and about 149 as a left-handed hitter at 2026 MLB rates. To get a share of one half for the observed split, a righty bat needs about 11 seasons and a lefty bat about 7. So an extreme split over a short stretch, especially a reverse split, mostly returns toward the average. In the "Tiny-sample reverse" example, the right-handed hitter has an OPS of .433 against lefties in 43 PA and .849 against righties, a gap of .416 toward right-handed pitchers. Because he has so few PA against lefties, only 1.9% of that gap counts, and the adjusted estimate points the other way: .035 toward left-handed pitchers, like an ordinary right-handed hitter.

The original work in The Book and at FanGraphs is built on wOBA. This calculator applies the same PA weights to the OPS gap, which you can compute from the record box alone, so treat the result as an approximation. It is meant to give a sense of how far to trust a split, not to forecast precisely.

When you copy numbers from a stats table

Most tables you will copy from share three quirks, and the box is built around them.

• No sacrifice-fly column: many split tables list AB, H, 2B, 3B, HR, BB and HBP but not SF. If you leave SF blank it counts as 0, so OBP comes out a little high. In the opening example, taking away the 2 sacrifice flies against lefties moves OBP from .394 to .399, up .005.
• Several seasons or several tables: type the counts as a sum, such as 312+64, in the cell. The calculator adds first and then recomputes all the rates from the totals.
• KBO's pitcher-type table: the KBO record room splits pitchers into three lines, left-handed, right-handed and underhand (the underhand line groups left-handed and right-handed submariners together, filter value LU,RU). So the KBO "right-handed" line leaves out submarine pitchers. To see a left-right split, add the underhand line to the side of the hand each underhand pitcher throws with.

The KBO record room shows these splits only in per-player tables, and no table of whole-league results by batting side could be found there. NPB's official site (npb.jp) did not load when checked on 2026-10-22, so whether it publishes left-right league totals could not be checked. The league ticks and the regression baseline in this calculator therefore use MLB values. When you read a KBO or NPB hitter, treat the direction and size as a rough reference, since the league environment is different.

Common mistakes

• Averaging two batting averages for the total: in the opening example, (.307 + .242) ÷ 2 = .2745, but adding hits and at-bats gives the real total of 138 ÷ 530 = .260. The average comes out higher because there are fewer at-bats against lefties.
• Dividing by plate appearances: the denominator for AVG and SLG is at-bats. Dividing by plate appearances, which include walks, hit-by-pitches and sacrifice flies, makes the value too low.
• Counting a hit-by-pitch as a walk too: if the same HBP goes into both the BB cell and the HBP cell, OBP is inflated. If your table lists combined "walks and HBP", split it before entering.
• Reading a few dozen PA of reverse split as skill: as the table above shows, with 100 PA against lefties only 4.3% of a right-handed hitter's observed split counts. An early-season "righty who struggles against lefties" mostly drifts back to the average.
• Reading a switch hitter's split the same way: a switch hitter's gap compares the right-side version of himself with the left-side version. That is why the page does not use the word "reverse split" for switch hitters and only compares the direction and size with the average.
• Leaving out the KBO underhand line: without it, the sample against right-handed pitchers shrinks, and the result shifts for a hitter who does especially well or badly against submariners.

Limits

• The 2,200 and 1,000 PA constants are the regression amounts The Book uses for wOBA splits. The calculator applies the same weights to OPS, so the result is an approximation. The 600 PA for switch hitters is not a published regression amount but the point where their split stabilizes, borrowed as a constant.
• The league ticks and the regression target are MLB 2026 regular-season values. The KBO record room shows pitcher-type splits only in per-player tables, so no league totals are included, and npb.jp did not load on 2026-10-22, so NPB could not be checked.
• The MLB side totals use each player's current registered batting side. A player who changed sides during the season is counted on one side only.
• The quality of the opposing pitchers is not adjusted for. FanGraphs gives the example of left-handed relievers who come in to face lefties and are often pulled before a good righty bats, so the mix of opponents can push split records away from true skill.
• Plate appearances are counted as AB + BB + HBP + SF. Sacrifice bunts and catcher's-interference reaches are left out, so the figure can be slightly lower than the official PA count.
• Each cell accepts 0 to 99,999. The comparison list lives briefly in your browser only; nothing is saved or sent to a server.

Sources

• 01MLB Glossary — On-base Percentage (OBP)Times on base counts hits, walks and hit-by-pitches but not errors or fielder's choices; sacrifice bunts are left out of the formula
• 02MLB Glossary — Slugging Percentage (SLG)Total bases divided by at-bats
• 03FanGraphs Library — Splits (Piper Slowinski, 2010-02-27)The platoon split (hitters do better against the opposite hand), the sample on the narrower side (75 to 100 PA against lefties in a 300-PA split), observed versus true splits, and the mix of hitters left-handed relievers face
• 04FanGraphs — Estimating Hitter Platoon Skill (Matt Klaassen, 2010-02-08)Regression per The Book: 1,000 PA for left-handed hitters and 2,200 PA for right-handed hitters of the league-average split against LHP; 2007 to 2009 wOBA splits of 8.6% (lefty bats) and 6.1% (righty bats)
• 05FanGraphs — Finding Switch-Hitters Who Should Stop Switch-Hitting (Jake Mailhot, 2022-01-11)According to the authors of The Book, a switch hitter's split stabilizes at about 600 PA against left-handed pitchers
• 06MLB Stats API — 2026 regular season hitter splits vs LHP and RHP (by player)1,308 player records added up by batSide (R, L, S); 2025 uses the same address with season=2025 (1,321 records)
• 07MLB Stats API — 2026 regular season team hitter splits vs LHP and RHP30-team total of 55,107 PA against LHP and 128,742 PA against RHP, matching the sum of the player records
• 08KBO record room — Hitter situational recordsSplits by pitcher type: left-handed, right-handed and underhand (left- and right-handed submariners together); no sacrifice-fly or plate-appearance column

Data as of 2026-10-22 The formulas and the regression constants were checked against the MLB Glossary and the original FanGraphs articles that day. The MLB averages by batting side were computed the same day from the MLB Stats API 2025 and 2026 regular-season records, and the layout of the KBO pitcher-type table was checked on the KBO record room the same day.

Frequently asked questions

Related tools