Start with the same 30-day performance record shown in the app’s Analysis screen. Below it, explore the separate Top Picks display ledger. These records cover different groups of picks, so their totals can differ.
Loading the app’s 30-day record…
Uses the same saved, graded picks and rolling 30-day game-time window as Analysis in the app. Archived picks and experimental shadow picks are excluded. This is a broader record than the Top Picks display ledger below.
Win rate = wins ÷ (wins + losses); pushes count toward total picks but not win rate. Benchmark units assume one unit risked per pick at −110 odds: a win earns about 0.91 units, a loss costs one unit, and a push returns the stake. Recorded returns apply the same one-unit stake at each pick’s saved odds; wins without a saved price are excluded, so the coverage count above shows how many returns are known. Neither figure is your personal betting returns. Data may take up to five minutes to refresh. Historical performance does not guarantee future results.
A narrower record of picks captured in the Top Picks display. Its 30-day window uses capture dates in Eastern time, rather than the rolling game-time window used by Analysis.
Loading the previous 30 days of official results…
Win rate, when shown, uses decisions only (wins + losses). Pushes, voids, and the pending count remain visible but do not enter that percentage. Pending pick details are never published here. Historical performance does not guarantee future results.
SharpScout measures the gap between the share of bets and the share of money on each side of a game. Every sport and market has its own starting points for Sharp Moderate, Sharp Good, and Sharp High, and those starting points can only move up as results come in. A gap has to clear the band for its sport before anything else is checked.
Freshness, line movement, available price, time to event, and source health have to support the read. A threshold by itself is not publishable evidence.
The analysis layer weighs supporting and conflicting evidence. If the case is incomplete, the correct decision is no pick.
Qualifying selections are timestamped into Top Picks. The tracked snapshot preserves what users saw before the result was known.
Signal means the current market data crossed a configured threshold. Candidate means an internal evaluation path considered the opportunity. Official Top Pick means the selection passed the publication gates and was shown to customers. The Top Picks display ledger counts only the third category. The App Analysis summary above covers the broader saved, graded record.
