How do I get PrizePicks data for backtesting?
Read the recorded line history, or the flat training set of graded lines, and reconstruct each line using only what was known at your decision time. The record includes one partial NBA capture on 27 Oct 2026, with broader recording from 28 Oct 2026: 215,827 lines and 149,536 moves so far, with gaps retained unless authentic timestamped observations can be recovered.
| League | Lines | Line moves | Graded |
|---|---|---|---|
| NBA | 80,754 | 58,747 | 45,368 |
| NFL | 23,315 | 40,551 | 7,453 |
| College Football | 19,123 | 23,404 | 6,494 |
| UCL | 7,994 | 12,096 | — |
| Tennis | 5,004 | 6,167 | 1,086 |
| EPL | 5,509 | 2,287 | 1,712 |
| La Liga | 4,507 | 1,513 | 1,741 |
| NHL | 22,241 | 1,178 | 9,269 |
| Soccer | 21,522 | 760 | 3,421 |
| College Basketball | 4,403 | 693 | 2,766 |
| Serie A | 6,745 | 466 | 1,816 |
| CS2 | 3,079 | 388 | — |
For each simulated decision, use the latest history entry observed at or before your cutoff; a later line must not leak into an earlier decision. Report the window, how many decisions you excluded for missing data, and your source_status, so a reader can tell the result from the sampling.
One line’s full history
curl "https://prizepicksapi.com/api/v1/projections/$PROJECTION_ID/history" \
-H "Authorization: Bearer $PRIZEPICKS_API_KEY"Go further
Related questions
Independent data service, not affiliated with PrizePicks. Information only, not betting advice.