Methodology · Calibration · Grading

How the record is built.

Every prop starts with a recorded line, confidence score, market, and direction, then gets graded against the final stat so the track record stays transparent.

Past performance does not guarantee future results. Propeller is a research and analysis tool, not a sportsbook.

Confidence by range.

Directional confidence counts strong OVER and strong UNDER signals the same way, then rolls each range into the full public record.

Loading lifetime confidence ranges...

The grading path

From model output to public result.

Each prop travels through the same path before it becomes part of the public track record.

01

Snapshot the recommendation

Player, market, line, direction, confidence, sport, game date, and book are recorded before the outcome is known.

02

Grade against the final stat

Overs need the final stat to finish above the line. Unders need it below the line. Exact ties are pushes.

03

Refresh the public record

Wins, losses, pushes, sport splits, and confidence ranges are recalculated from the public results API that powers this record.

What the record means.

The goal is not to make the number sound bigger. The goal is to make the number understandable.

Pushes excluded

Pushes are shown in the record, but they are tracked separately from wins and losses.

Sample size matters

A large all-time record gives context, while confidence ranges show how different signal strengths have performed.

Confidence is not certainty

A confidence score is a research signal. It is not a guarantee, and individual outcomes still carry variance.

Results are inspectable

The live results page shows recent graded props so users can see the data path instead of only a headline number.

Founder 500

Use the record, then research today's board.

Founder 500 members get free lifetime core access while launch spots remain.

Get Free Lifetime Access