"Proj" updates live: once a player has actually scored this week, their real points replace the pre-game estimate, converging to the true final total as games finish — not a number frozen at kickoff. Win % converts the gap into a probability using the normal distribution and your league's real week-to-week scoring volatility. "Yet to play" shows how many starters (and at which positions) haven't recorded a live score yet — how much of the matchup is still actually in motion.

Starters

Bench

Flags a bench player as a likely upgrade using a projection blending each player's own scoring history — this season's games once they exist, or that exact player's own last-season average before then — with a defense-vs-position model built from real box scores. Only a player with zero track record anywhere (a true rookie) falls back to a generic position average. A starter's own injury designation also adjusts their number: Questionable/Doubtful applies a real risk discount (15%/50%) rather than an automatic swap, and Out/IR/Suspended zeroes it, since Sleeper itself expects zero snaps. No expert rankings or opinions involved. The status line below reports a leaguewide custom-scoring match rate (spans every NFL player, so it's diluted by backups/inactives with no stats to match) alongside a rostered-players-only rate — the one that actually reflects whether your league's real players are being scored correctly.

Every available (unrostered) player, ranked by the same projection model used in Start/Sit — filter by position like Sleeper's own Available Players screen, or leave on All to browse everything at once. A card is flagged when it fills a thin roster spot or would outscore your current weakest player there. Informational only — Sleeper's API is read-only, so make the actual claim in the Sleeper app.

Player trade values from FantasyCalc, a community market-value tool that keys directly off Sleeper player IDs. Pick players on each side to compare total value.

Ranks every team by total projected points across their actual current starters this week — the same projection model used in Start/Sit (recent scoring blended with real defense-vs-position data). This is a strength ranking, not your actual W-L standings, and a team with an easy schedule of blowout wins can rank lower here than one that's been in close, competitive games.