How 20-0 Works
The 20-game gauntlet
Every run is the same path: 17 regular-season games, then the playoffs: the Divisional round, the Championship game, and the title game. That is 20 wins, not 21, because winning out takes the 1 seed and the 1 seed earns the bye. The road gets heavier the deeper you go, and the last three are all heavyweights.
Franchise-era spins
Every pick spins up a franchise and a decade (the '70s Steelers, the '90s Cowboys) and puts that era's whole roster on the board, every position, top to bottom. You get one new team and one new era per draft. Take the best player on the board, or the one your roster actually needs.
The Impact XI
Eleven picks, both sides of the ball: quarterback, running back, two receivers, tight end, an offensive line anchor, then an edge rusher, interior lineman, linebacker, corner, and safety. You build the offense first; the defense unlocks once it is set. Skill Six is the sprint: offense only, with the defense set to league average. Two-Platoon adds a four-man bench.
The value backbone
Offensive linemen do not have box scores. That is the point. 20-0 leans on a season-wide value signal, surfaced as a 0–100 position rating and a line grade, never a raw number, so a great left tackle counts like the great left tackle he was, not like a zero. The same backbone steadies early-era defense, where the stat sheet runs thin.
Hard caps
No quarterback, no parade. A roster without a real passer has a ceiling, no matter who is catching. Same story with no pass rush and no coverage anchor: the model carries hard caps for the things January punishes, and stacking five receivers will not buy your way around them.
Era-normalized, position by position
Every player-season is scored as a percentile within its own season and position group, so a 1970s workhorse back and a 2010s slot receiver go head to head on the same scale. Raw counting stats never compare across eras; percentiles do.
Deterministic, shareable, ranked
The same roster always produces the same result, and every result is locked in exactly as it was scored, so old share pages hold up. Leaderboards are per mode; scouting level changes what you see while drafting, never the outcome. A 20-0 result is a calibrated projection, not a claim. It starts the argument, it does not settle it.