Power rating blends each roster's projected optimal lineup with what it has actually scored, weighted toward real results as the season goes — right now one game in, it is roughly one third results. Luck compares wins to expected wins: how often that team's weekly score would have beaten a random opponent. A team at −0.8 has been beaten by the schedule, not by the league. Matchup projections use each manager's currently set lineup, not their best possible one, so a bad start shows up as a lower number.

Implied total is what the betting market expects each team to score: the game total split by the spread. It is the single best one-number read on how much offense a game should produce. Matchup chips grade the defense each offense is facing by fantasy points allowed — and with one week played that is one game of evidence, so treat it as a tiebreaker, not a reason to bench a starter.

Where this comes from. The headlines are RotoWire's public feed, read through our own API and cached ten minutes, so a story lands here within minutes of them posting it. The injury board underneath is the designation ESPN and Sleeper carry for every player on the board — terser, but complete, and it is what actually moves a projection. A player marked out is valued at zero for the week; questionable takes an eight percent haircut, injured reserve forty-five.

Both tools answer the same question: what would a best legal starting nine score? Adding up player values calls most trades a wash for both sides; running the same players through each lineup shows who they actually displace. Where the gap is smaller than the model can resolve — 1.5 points a week, 0.8 rest of season — it says so instead of guessing.