All of baseball, in one place.
A personal data lake behind every project on this site: every game since 1871, every tracked pitch, every roster move, and my own swings and guesses, joined on one set of ids. This page is the plan and the catalog. Most of it isn't built yet, and each dataset says so.
- Datasets
- 41
- On hand
- 8
- Rows on hand
- 3.9M
- Seasons on hand
- 1901–2026
- Ideas ready now
- 2 of 23
Roadmap
Ordered by the ideas each phase makes buildable- P0Now
Foundations
Storage, the catalog, and every existing cache moved into the lake.
Game results · Standings · Relief pitching · Best stretches · Every pitch · The show's questions · Read the Swing plays · Read the Swing guesses
- P1Up next
The spine
One id for every person, team, park and game since 1871, plus season stats and game logs.
People · Teams and franchises · Ballparks · The game key · Rules timeline · Game logs · Player seasons
Ideas it makes ready98 on-air asks behind them - P2Planned
Every pitch
Pitch tracking back to 2008 with every column, plus fielding, running, umpires and video links.
Pitch flight · Fielding · Running game · Umpires · Video index
- P3Planned
Every play
Play-by-play and box scores across history, full game feeds, the Negro Leagues, weather and travel.
Box scores · Play-by-play · Negro Leagues · Full game feeds · Weather · Rest and travel
- P4Planned
Front office
Transactions, the injured list, rosters, the draft, contracts and payroll.
Transactions · Injured list · Daily rosters · The draft · Contracts · Payroll
- P5Planned
The pipeline
The minors, other leagues and amateurs.
Minor leagues · Triple-A tracking · Other leagues · College and amateur
No idea waits only on this yet - P6Planned
Personal
Cage sessions, games I've been to, a ledger of takes, fantasy history and gear.
Cage sessions · Games I've been to · Take ledger · Fantasy league · Gear specs
Ideas it makes ready - P7Planned
Open it up
A query console on this page for the tables whose terms allow sharing.
A site feature, no new data
Coverage
Seasons each dataset covers, 1871–2026Catalog
41 datasets in 8 domainsIdentity
The spine: who, where and when, on one set of ids.
People
P1 · The spineSpineEvery player, manager, coach and umpire, with the ids MLB, Retrosheet, Baseball-Reference and FanGraphs use for them, so any two sources can be joined.
FeedsEvery dataset joins through itTeams and franchises
P1 · The spineSpineEvery team-season tied to its franchise, league and division, Negro Leagues included, so the 1901 Milwaukee Brewers and the 2026 Orioles line up as one franchise.
FeedsDivision GapBallparks
P1 · The spineSpineEvery park with its location, elevation, roof and fence distances, season by season as the fences move.
The game key
P1 · The spineSpineOne id per game that links Retrosheet's game ids to MLB's, so a game from 1923 and a game from last night join the same way.
FeedsEvery dataset joins through itRules timeline
P1 · The spineEvery rule change that moved the numbers, with its effective date: the lower mound in 1969, the DH, the extra-innings runner, the 2023 pitch clock, shift limits and bigger bases, and the 2026 ABS challenges.
FeedsEra Dial
History
Games, seasons and plays back to 1871.
Game results
On handEvery regular-season game since 1901: date, teams and final score, checked against the official standings.
Standings
On handOfficial final standings for every season since 1903.
Relief pitching
On handEach team's bullpen totals for every season since 1974.
Best stretches
On handEvery team-season's best 30, 36 and 50-game runs, built from the game results.
Published on this site as Claim Check's data file, the lake's first published table.
Game logs
P1 · The spineEvery major league game since 1871 with starting lineups, umpires, attendance, game length and the line score.
Retrosheet publishes each season after it ends.
Player seasons
P1 · The spineBatting, pitching and fielding lines for every player-season, plus awards and the Hall of Fame.
Box scores
P3 · Every playEvery player's line in every game: the source for game logs, pitch counts and rest.
Pitcher lines for 2018–24 are already on disk from an older project.
Play-by-play
P3 · Every playEvery plate appearance with the count, runners, outs and result, for as much of history as was scored.
Coverage follows Retrosheet, complete for recent decades and thinner in the early years.
Negro Leagues
P3 · Every playThe Negro Leagues seasons MLB recognizes as major league, with games, players and stats on the same ids.
FeedsClaim Check
Tracking
Pitch, swing, fielding and running measurements from the tracking era.
Every pitch
On handPitch-level Statcast: speed, spin, movement, location, the count and the result, plus bat speed and swing length since mid-2023.
Phase 2 backfills to 2008, when pitch tracking began, and keeps every column Savant offers.
Pitch flight
P2 · Every pitchThe full measured flight of every pitch: release point, velocity and acceleration in three dimensions, enough to redraw any pitch's path.
FeedsTunnel VisionFielding
P2 · Every pitchOuts above average, catch probability, outfield arm strength, and catcher framing, blocking and pop time.
Other defensive systems have stricter terms, so Glove Consensus may only be able to link to them.
Running game
P2 · Every pitchSprint speed, runners' leads and jumps, pitchers' times to the plate, and baserunning runs.
Full game feeds
P3 · Every playMLB's complete record of each game, with every play, pitch, fielder, substitution and umpire, kept exactly as published.
Kept raw; the play-by-play and box score tables are built from it.
FeedsSteal Lab
Front office
Rosters, transactions, contracts and money.
Transactions
P4 · Front officeTrades, signings, releases, call-ups and options, each with its date.
Older transactions are less complete.
FeedsPlayer FactoryInjured list
P4 · Front officeEvery injured-list stint, from placement to return, with the injury where it was reported.
FeedsVelo WatchDaily rosters
P4 · Front officeWho was on every active roster on every day of the season.
FeedsLineup LabThe draft
P4 · Front officeEvery pick since the first draft in 1965, and what became of each player.
FeedsPlayer FactoryContracts
P4 · Front officeContract terms and payment schedules, including deferred money, options and opt-outs, starting with the largest active deals.
FeedsContract X-RayPayroll
P4 · Front officeEach team's opening-day payroll by season.
Lahman has player salaries through 2016; later seasons are compiled by hand.
Development
The minors, amateurs and other leagues.
Minor leagues
P5 · The pipelineGames and season stats for every affiliated level, from Triple-A down to the complex leagues.
FeedsPlayer FactoryTriple-A tracking
P5 · The pipelinePitch-level tracking for Triple-A, the last test before the majors.
Savant publishes it for recent seasons only.
FeedsBreakout ClockOther leagues
P5 · The pipelineSeason stats from the top leagues in Japan, Korea and Taiwan, plus the World Baseball Classic, for translating numbers between leagues.
FeedsPlayer FactoryCollege and amateur
P5 · The pipelineCollege and summer-league season stats, for players before the draft.
FeedsNo project yet
Environment
Weather, umpires and travel.
Weather
P3 · Every playFirst-pitch weather for every game, and hourly temperature, wind and humidity from the weather station nearest each park.
NOAA records are public domain; station coverage thins before 1973.
Umpires
P2 · Every pitchWho worked every game, and the called zone of each plate umpire, measured pitch by pitch.
Called zones start in 2008, with pitch tracking.
FeedsZone ShiftRest and travel
P3 · Every playMiles traveled, time zones crossed and days of rest before every game, worked out from the schedule and park locations.
FeedsVelo Watch
Media
The show's questions, game clips and video links.
The show's questions
On handEvery question the Talkin' Baseball hosts asked on air, transcribed and sorted, grouped into the 44 questions on the board.
Transcripts stay private. The question board on this site is the published part.
FeedsAnswers for CoachRead the Swing plays
On handThe plays behind Read the Swing, each with its Statcast record and a trimmed clip.
Video index
P2 · Every pitchA link from every tracked pitch to its public video on Baseball Savant, so any study can pull clips.
Links only. No video is stored.
FeedsRead the Pitch
Personal
Data I collect myself.
Read the Swing guesses
On handEvery guess made in the game: the play, how early the video froze, the call and how long it took. Visitors' guesses are anonymous.
FeedsRead the SwingCage sessions
P6 · PersonalEvery ball hit in front of the homemade tracker: exit speed, launch angle, spray, spin and a slow-motion clip.
Other hitters' sessions stay private; only my own swings could be published.
FeedsCage CamGames I've been to
P6 · PersonalA ledger of every game I've attended, joined to its box score and play-by-play, so each ticket opens the whole game.
FeedsField NotesTake ledger
P6 · PersonalPredictions and claims, mine and ones heard on the show, logged with a date and graded when the answer comes in.
FeedsClaim CheckFantasy league
P6 · PersonalDrafts, rosters, trades and matchups from my fantasy league, season by season.
League-mates' names stay private.
FeedsNo project yetGear specs
P6 · PersonalBats and gloves measured: length, weight, balance point, pocket depth and break-in, with notes over a season.
Solid: the project needs it. Faint: the project could use it later.
How it's built
Parquet files, queried with DuckDB- Surface
Published
Small extracts this site serves, cleared against each source's terms. Claim Check's streaks file is the first.
- −1
Marts
Tables ready for analysis: games, plays, pitches, player seasons, rosters.
- −2
Clean
Typed, de-duplicated and joined to the lake's ids for people, teams, parks and games.
- Floor
Raw
Every file exactly as downloaded, never edited, with where and when it came from.
- Keep every raw file. Any table can be rebuilt from what was downloaded.
- One id for everything. Each person, team, park and game gets one lake id, mapped to every source's ids.
- Every number traces back. Each table records the files and steps that built it.
- Check before publishing. Rows against the schedule, totals against official records, keys that must be unique.
- Terms decide what's public. Open data can be shared. MLB data is for personal use, so the site shows analysis, not copies. Private data stays in the lake.
- Refresh on a schedule. Daily in season for pitches, games and transactions; yearly for the historical releases.
Terms
What each label in the catalog means- Open data 7
- Openly licensed. Can be shared with credit.
- Personal use 20
- MLB data, for individual, non-commercial use. The site shows analysis built on it, not copies of it.
- Hand-built 4
- Compiled by hand from public sources, with each fact cited.
- Mine 4
- Data I create.
- Private 3
- Kept off the site.
- Terms to confirm 3
- Not used until the source's terms are checked.
MLB data is used under MLB Advanced Media's terms, which allow individual, non-commercial use. When Retrosheet data arrives, its required notice will appear here and on every page that uses it.