Loading ticker...
Loading… @TheOddsGap

The Odds Gap Data

Sports Odds Archive

Every scan we captured since June 4, 2026, one file per day, nothing backfilled or smoothed, so a backtest sees the prices that actually existed then.

123 days on file since June 4, 2026. Last day added October 4, 2026.

What's in It

  1. Every scan of game lines since June 4, 2026, one file per UTC day.
  2. Player props from August 4, 2026, at the same four checkpoints.
  3. The same columns as the recent export, plus a marks column.
  4. Known issues named on every row they touch, with a notes file.
  5. 123 days on file, each ready 24 hours after it ends.

Download

The last 7 days, free. No signup, no key.

curl -o last-7-days.csv "https://theoddsgap.com/api/odds-export.csv?sport=all"

Try it: /data/sample.csv holds every book's prices on one finished NFL game, New York Jets at Chicago Bears. Free, no signup, no key.

Known issues: ProphetX prices before the August 26, 2026 repair and Novig sizes before September 28, 2026 are marked on the rows they touch. Every issue by date is in the notes file.

Days on file: 123, from June 4, 2026 to October 4, 2026. The list of days is at /api/archive/days.

Questions

Can I use the archive for backtesting?

That is what it is for. Each file holds prices as they stood at that scan, with nothing backfilled or smoothed, so a backtest sees only what was on the board then. It comes with prices, never picks.

Why does the archive start on June 4, 2026?

That is the first day we kept every scan. Nothing earlier exists anywhere, so we do not sell it. The archive grows by one file every day.

Is the archive free?

No. The archive is part of the subscription and needs your key. The last 7 days of the regular export stay free, and so does the sample file at /data/sample.csv: one finished NFL game, New York Jets at Chicago Bears, from every book.