Play-by-Play Events

Retrosheet event files (.EVA, .EVN) record every pitch, play, substitution, and baserunning event in Major League Baseball.

retrosheetpy parses these event archives into a 164-column strongly-typed schema via chadwickpy.

Usage

import retrosheetpy as rs

# Download & parse 2024 plays (returns Polars DataFrame)
events = rs.events(year=2024, backend="polars")

# Filter for home runs
home_runs = events.filter(events["HR_FL"] == True)
print(f"Total 2024 MLB Home Runs: {home_runs.height}")

Key Column Reference (164 Columns)

The full 164 columns are categorized below:

1. Game & State Identifiers

Column	Type	Description
`GAME_ID`	`string`	Retrosheet game identifier (e.g. `BOS202404090`)
`YEAR`	`int32`	Season year (e.g. `2024`)
`INN_CT`	`int16`	Inning number (1–9+)
`BAT_HOME_ID`	`int16`	Batting team flag (`0` = Visiting, `1` = Home)
`OUTS_CT`	`int16`	Outs before play occurred (0, 1, 2)
`BALLS_CT`	`int16`	Balls before play occurred (0–3)
`STRIKES_CT`	`int16`	Strikes before play occurred (0–2)
`PITCH_SEQ_TX`	`string`	Pitches thrown in the plate appearance (e.g. `BBSFX`)

2. Player Identifiers

Column	Type	Description
`BAT_ID`	`string`	Retrosheet player ID of batter (e.g. `ohtas001`)
`BAT_HAND_CD`	`string`	Batter handedness (`L`, `R`, `S`)
`RESP_BAT_ID`	`string`	Batter responsible for the plate appearance
`PIT_ID`	`string`	Pitcher ID
`PIT_HAND_CD`	`string`	Pitcher handedness (`L`, `R`)
`RESP_PIT_ID`	`string`	Pitcher responsible for runners / run
`POS2_FLD_ID` - `POS9_FLD_ID`	`string`	Player IDs for fielders by position (Catcher through RF)

3. Base Runners

Column	Type	Description
`BASE1_RUN_ID`	`string`	Runner on first before play (empty if vacant)
`BASE2_RUN_ID`	`string`	Runner on second before play
`BASE3_RUN_ID`	`string`	Runner on third before play

4. Play Details & Scores

Column	Type	Description
`EVENT_TX`	`string`	Raw Retrosheet play descriptor (e.g. `HR/7/F`, `K`, `S8/L.1-3`)
`EVENT_CD`	`int16`	Chadwick numeric event code (e.g. `23` = Home run)
`AWAY_SCORE_CT`	`int16`	Visiting team score before play
`HOME_SCORE_CT`	`int16`	Home team score before play
`RBI_CT`	`int16`	Runs batted in on play
`EVENT_OUTS_CT`	`int16`	Number of outs recorded on play (0, 1, 2, 3)

5. Boolean Indicators

Column	Type	Description
`AB_FL`	`bool`	At-bat indicator
`H_FL`	`bool`	Hit indicator (single, double, triple, HR)
`SH_FL`	`bool`	Sacrifice hit / bunt
`SF_FL`	`bool`	Sacrifice fly
`DP_FL`	`bool`	Double play turned on play
`TP_FL`	`bool`	Triple play turned on play
`HR_FL`	`bool`	Home run hit on play
`WP_FL`	`bool`	Wild pitch occurred
`PB_FL`	`bool`	Passed ball occurred

Multi-Season Processing

You can fetch multiple seasons in a single call. retrosheetpy automatically concatenates the Arrow record batches efficiently:

import retrosheetpy as rs

# 2020 through 2024 (approx 900,000 plays)
multi_events = rs.events(year=range(2020, 2025), backend="polars")