MLB
67.5%
Historical analysis archive
Read the unit before the rate.
This archive separates raw graded analysis rows from entries retained under the current public API collapse rules. It includes repeated snapshots and legacy retrospective data, so it is not a uniquely published forward-test or ROI record.
By Scott Olmer, Founder · Reviewed July 13, 2026
Past performance does not guarantee future results. Pushes are tracked separately from wins and losses.
Raw rows by sport.
These sport totals count raw historical analysis rows. Repeated books, lines, snapshots, runs, and retrospective records may be represented.
NHL
62.3%
NHL
NBA
58.9%
NBA
MLS
85.0%
MLS
EPL
83.0%
EPL
NFL
59.8%
NFL
Confidence ranges
Signal strength, reconciled.
Confidence is directional: a strong UNDER and a strong OVER are both high-confidence model signals. These descriptive ranges reconcile to the raw historical database, not the collapsed ledger.
High Confidence
75+143K raw rows · 88.9% UNDER · 56.1% OVER
Medium Confidence
65-74712K raw rows · 79.2% UNDER · 59.3% OVER
Slight Confidence
55-641.07M raw rows · 65.5% UNDER · 55.5% OVER
Recent ledger sample.
A sample of rows retained after the current API collapse rules, including line, signal score, actual result, and book. This feed does not prove pre-event publication.
| Date | Sport | Player | Market | Line | Conf | Actual | Result | Book |
|---|---|---|---|---|---|---|---|---|
| 2026-07-30 | mlb | Nasim Nuñez | Home Runs UNDER | 0.5 | 80.0% | 0.0 | WIN | Prizepicks |
| 2026-07-30 | mlb | Nick Gonzales | Stolen Bases UNDER | 0.5 | 80.0% | 0.0 | WIN | Prizepicks |
| 2026-07-30 | mlb | Edgar Quero | Doubles UNDER | 0.5 | 80.0% | 0.0 | WIN | Betr |
| 2026-07-30 | mlb | Ronald Acuña Jr. | Total Bases UNDER | 1.5 | 80.0% | 4.0 | LOSS | Betr |
| 2026-07-30 | mlb | Ryan McMahon | Stolen Bases UNDER | 0.5 | 80.0% | 0.0 | WIN | Betr |
| 2026-07-30 | mlb | Paul Goldschmidt | Stolen Bases UNDER | 0.5 | 79.9% | 0.0 | WIN | Betr |
| 2026-07-30 | mlb | Ozzie Albies | Stolen Bases UNDER | 0.5 | 79.8% | 0.0 | WIN | Prizepicks |
| 2026-07-30 | mlb | Spencer Jones | Stolen Bases UNDER | 0.5 | 79.8% | 1.0 | LOSS | Betr |
| 2026-07-30 | mlb | Edgar Quero | Stolen Bases UNDER | 0.5 | 79.8% | 0.0 | WIN | Prizepicks |
| 2026-07-30 | mlb | Ian Happ | Stolen Bases UNDER | 0.5 | 79.7% | 0.0 | WIN | Underdog |
How grading works
The grading rule is simple; the archive needs context.
Rows are graded after games finish by comparing the stored side and line with the final stat. Pushes remain separate. Publication time, model version, evaluation mode, and consistent price data are not available on every historical row.
01
Analysis row stored
The historical feed exposes player, market, line, direction, signal score, sport, game date, and book. It does not establish pre-event publication for every row.
02
Outcome graded
Final stat values are compared to the line. Overs win above the line, unders win below the line, and exact ties are pushes.
03
Record updated
Wins, losses, pushes, totals, and sport splits refresh from the public results API.
Limitations and definitions
What you should know before quoting this archive.
The historical database and collapsed ledger are research records. Neither is a clean cohort of uniquely published recommendations. See how grading works and the editorial and corrections policy.
01
Two different units
The summary and confidence endpoints count raw graded analysis rows. The public ledger applies a collapse key and therefore reports a smaller total.
02
Historical data is mixed
Repeated snapshots and legacy retrospective analysis are included. The current feed does not label every row as forward-tested or backtested.
03
No ROI claim
Price and payout are not consistently present, so the descriptive outcome rate cannot support a defensible profit or return-on-investment claim.
Founder 500
Free lifetime core access while launch spots remain.
Review the record, then use the web app to research today's board with the same model infrastructure behind this ledger.
How are Propeller Picks results graded?
Results are graded against final box score values after games complete. Pushes are tracked separately from wins and losses.
Does the historical archive guarantee future results?
No. The archive includes repeated analysis snapshots and legacy retrospective data. It is research context, not a forward-tested ROI record, and past performance does not guarantee future results.