17-0 is a free fantasy-football roster game — the gridiron sibling of the viral basketball game 82-0. You build an all-era fantasy lineup from over fifty years of NFL history and find out whether it's good enough to win every game and finish a perfect 17-0 season.
A flawless NFL regular season is 17 wins and zero losses — only the 1972 Dolphins (then 14-0) and the 2007 Patriots (16-0) ever came close. Your job is to draft a 9-player fantasy roster deep enough that the season simulator gives you all 17. It's deliberately hard: only a near-perfect roster gets there, so a true 17-0 is the holy grail.
A standard fantasy lineup:
Pick a mode on the home screen before you draft. Whichever you choose, the mode is shown above the reels the whole time so you always know which game you're playing.
The slot machine spins you a random franchise and decade. You draft one player-season from that team and era to fill an open slot, then do it eight more times. Full stats are shown on every card.
The same slot-machine draft, but the numbers are hidden — you pick on football knowledge and instinct alone. The cards reveal who they are, not how they performed.
The slot machine plus a $250 million cap. Every player now carries a price, and here's the catch: prices are based on raw yardage and gaudy box-score volume — the stuff that gets players paid. The simulator, meanwhile, scores within-year fantasy value. So a high-volume, low-efficiency compiler is expensive and ordinary (a trap), while a quietly elite season is a bargain. Hunting that gap under the cap is the whole skill.
The whole game flipped upside down: draft the worst roster you can and try to lose all 17. The reels still spin — now you're hunting scrubs and camp bodies, and your best player is the ringer who keeps accidentally winning you games. A perfect 0–17 (à la the 2017 Browns) is its own kind of holy grail.
Salary Cap's evil twin. You've got the full $250M and the goal is to waste every dollar — overpay for gaudy-yardage busts who can't actually move the needle. You're judged on dollars per win: spend the most, win the least. Because price tracks value in football, you won't go winless — and that's the joke: the most expensive mediocrity in football.
One shared puzzle for everyone, every day — a special twist rule (for example, "every player from a different franchise," "one team only," or a single decade). You get one ranked attempt, and your result lands on a global leaderboard; come back the next day to keep your streak alive. Your standing shows on the home screen whenever you reopen the app.
Football solved the "compare across eras" problem with fantasy points, and 17-0 grades every player on his percentile within his own season. That means a 1975 running back is judged against 1975's backs and a 2023 one against 2023's — so the passing-era explosion washes out and a great year is a great year, whatever the decade. Team defenses are scored as a single D/ST unit (sacks, takeaways, defensive & return TDs, points allowed), which is why noisy individual defenders never clutter the board.
Once all 9 spots are filled, your roster plays a 17-game schedule. The engine turns your lineup's mean within-year percentile into a win expectation: a league-median roster lands around 9-8, and only a roster that's elite at every slot brushes 17-0. A single soft spot drags the whole season down — a perfect team has no weak links.
Player stats cover completed NFL regular seasons from 1970 through 2024 across the 32 current franchises (relocations are merged, so the Titans include the Houston Oilers and the Cardinals include St. Louis). The modern era (1999+) comes from nflverse; the 1970–1998 seasons are reconstructed from Pro-Football-Reference mirrors (offense), footballdb (kickers, with field goals broken out by distance), and team-defense data — all scored on the same within-year fantasy percentile so every era is fair.
Play 17-0 →