MLB Head-to-Head Records Since 2002

All 435 normalized matchups covering 59,834 meetings since 2002. Those current-feed totals power recent form and rate cards. 20 reviewed matchups also separate out Retrosheet history through 2025.

435
Normalized matchups
59,834
Current-feed games
28.3%
One-run share
9 runs
Average total

Data updated

Start with the closest pairingSeattle Mariners vs Tampa Bay Rays46.6%

Published matchups

Choose a league, then open the directory to jump to any published record.

All matchups435 matchup pagesOpen directory

Interleague pairings meet in short bursts, so their percentages rest on far smaller samples than a divisional pairing. Each page flags its own sample size before quoting a rate as meaningful.

MLB head-to-head FAQ

What MLB matchups does this hub cover?

This hub publishes all 435 normalized pairings among the 30 current MLB clubs. Every matchup uses completed final scores from the current feed starting in 2002. 20 reviewed matchups also include a separate Retrosheet history window and reviewed query-owner titles. The remaining 415 use the standard deterministic title and H1 template.

Why do some matchup pages show two date ranges?

20 reviewed matchups combine Retrosheet game logs through 2025 with later final games from the continuously refreshed feed. Recent games, rate cards and betting context always use the current feed from 2002. The other 415 pages use that feed alone. Each page states its source window so the record and narrower analysis cannot be confused.

Do these pages show results against the betting line?

No, and we say so on every page. We hold no historical sportsbook prices for these games, so nothing here is an against-the-spread record or a closing-line result. What you get instead is outcome frequency: how often a pairing stayed inside one run, ended with a margin of two or more, or crossed a half-run total threshold. Those are facts about final scores rather than claims about old prices.

How should a bettor actually use a head-to-head page?

Treat it as historical context, not a prediction. Start with the series record, then compare the one-run share and scoring average with the recent 20-game sample. A visible difference shows how recent finals diverge from the longer archive. Starting pitchers, lineups, bullpens, weather and venue still matter more to the next game than this record does.

How are MLB doubleheaders counted?

Each completed game counts once, including both games of a doubleheader when the source gives them separate game IDs. The 435 published pairings currently contain 59,834 meetings from 2002 onward. Spring training and games without a final score are excluded. Score or schedule anomalies are cross-checked against the official MLB Stats API before the matchup files are rebuilt.

When were these MLB records last updated?

The most recent published matchup data changed on September 5, 2026. This hub shows the newest data-change date across all 435 pairings, while each matchup keeps its own update date. A source check with no new or corrected final does not advance any public date. If a listed score or date looks wrong, send the matchup and game date through the contact page for source verification.

Other sports

The same treatment across the rest of the head-to-head section.

Methodology & Data Sources

Current MLB game logs come from the balldontlie API and cover the 2002 season onward. Scores and schedule anomalies are checked against the official MLB Stats API. Spring training is excluded, while results from earlier franchise identities are merged into the current club.

The History Lab uses Retrosheet game logs through 2025, with inning lines, team totals, attendance, duration and ballpark fields when available. Its source window is labeled separately from the refreshed current feed that starts in 2002.

The betting context summarizes completed final scores only. It contains no sportsbook closing lines and summarizes only completed games. Report a possible error through our contact page with the matchup and game date.

Read the full H2H data methodology →

Last updated Sources: balldontlie, MLB Stats API, Retrosheet