Qualification comes before rank
Short windows require at least 10 graded decisions. All-time, sport, and market boards require at least 25 graded decisions. CLV rankings require at least 25 valid closing captures. A board with no qualifying rows is omitted instead of publishing an empty or misleading page.
Only normal accounts and currently eligible public records can participate. Demo, internal, private, delayed-unrevealed, deleted, or restricted data is not projected into public rankings.
Metrics answer different questions
Net units measures accumulated graded profit. ROI divides net profit by graded risk. Average CLV summarizes qualified closing-price comparisons. The Most Consistent board uses the documented LBCI formula, which combines a 95% Wilson lower-bound win rate with confirmed-odds break-even context and a sample floor.
No one metric proves prediction skill. A rank can change when records are graded, corrected, become public or private, or move into or out of the stated period.
How to evaluate a row
Open the linked public profile and review sample size, wins, losses, pushes, average odds, risk volume, net units, ROI, CLV count, and recent receipts. A large profit total can reflect more risk, while a high ROI from a small sample can be unstable.
Compare records only under the same metric and period. Do not treat a seven-day result as equivalent to an all-time record.
Updates, corrections, and limitations
Each detail page states its period, qualification, tie-break order, generation time, and backend methodology. Source scores, grades, odds, and closing captures can be delayed or corrected. The page reflects the latest qualified public representation rather than preserving an immutable ranking snapshot.
Leaderboards describe past public records. They do not guarantee future performance and are not recommendations to follow a user or selection.
What PickProof deliberately does not publish
PickProof does not create a leaderboard page merely because a slug exists. NFL remains absent while it has no qualified rows, and Most Improved remains omitted because no approved reproducible methodology is published. Demo rankings, screenshot-only claims, private profiles, empty profiles, and underqualified records are not used to make the public footprint look larger.
Historical archives will be added only when a frozen period, eligibility policy, and durable source record can be reproduced. Until then, the available pages represent current qualified views rather than pretending to be permanent historical research.
Leaderboards use eligible public PickProof records only. Past results, ROI, CLV, rank, and badges do not guarantee future performance and are not betting advice.
Published and maintained by PickProof, LLC. Last reviewed 2026-07-30. PickProof reviews this material against the product's implemented recordkeeping rules, public-data contract, and disclosed calculator formulas.
