How the 20-0 Engine Works

A game that asks you to chase a complete perfect season — 17 regular-season wins and three playoff wins, Super Bowl included — owes you an honest look at how it decides every one of those games. No black box. Here is exactly what happens under the hood each time you sim a 20-0 season, and then a real single-elimination bracket, from the postseason back.

The one-line model

Every game resolves to a net rating — a single number that says how many points better (or worse) your roster projects to be that week — which the engine then converts into a final score with a controlled amount of randomness. Positive net rating, you're favored; the bigger it is, the safer the win. The score itself is built off a scoring baseline (tuned by the era's pace), with your margin set by net rating divided down and a little extra game-level noise on top. Everything below is just the honest list of what feeds that number.

What goes into net rating

Your offense vs. their defense, and their offense vs. your defense. The core matchup, and the largest term. A great offense against a soft defense opens up a lead; a great defense caps what the other side can do.
The trenches. A separate line-of-scrimmage term — blocking and run defense against the opponent's front — nudges the rating on top of the raw offense/defense diff. Games are still won up front.
Home-field edge. A flat swing for playing at home versus on the road — the same few points the real league has lived with for decades. In the bracket, a strong seed hosts more often, so your regular season is quietly buying you playoff home games.
Fatigue. The deeper into the season you go, the more a low-stamina roster wears down — the penalty scales with how far along you are and how thin your conditioning is. Perfect seasons are lost in November and December, not September, so late weeks get harder on purpose.
Injury risk. Each week carries a small chance a durability-thin roster takes a hit that drags down that game's rating. It's rare, it's luck, and it's exactly the kind of thing that ends real unbeaten runs.
Seeded variance. Football isn't deterministic, so we add a random swing to every game from a seeded generator — the same roster and season replays identically, but any given week can still break either way.

Before the games: chemistry and era fit

Your nine drafted legends aren't just averaged together. Before a single snap, the engine scores the roster's chemistry with named, transparent bonuses and penalties — an Elite Passing Attack, Dominant Trenches, a Shutdown Secondary, a Relentless Pass Rush, and so on, against holes like Leaky Protection or a One-Dimensional Offense. It also scores era fit: legends drafted for the era you spun play up; stylistic misfits play down. Those two numbers fold into the team ratings that then drive every net rating above, so a well-built, era-appropriate roster is favored more often — which, across 20 games, is the whole ballgame.

The part most competitors skip: a real playoff bracket

An undefeated regular season only earns you a seed. 20-0 then simulates a genuine single-elimination bracket in the NFL's real format. Go 17-0 and you lock the #1 seed and its bye, which means a complete run is exactly three more wins — Divisional, Conference Championship, Super Bowl. A strong seed earns a modest home-field bump and hosts more rounds, but it's deliberately modest: one bad game, at any point, ends the run with no next week to fix it. That is why the site is named for 20 wins and not 17. Seventeen buys you the best seat in the tournament; only twenty finishes the job.

From rating to a final score

The net rating is centered on a realistic scoring baseline, split between the two teams, and nudged by a little more game-level noise so scores land where football scores actually land. If the two sides tie, we break it — there are no ties on the road to 20-0. The result is a win or a loss, and the season, and then the bracket, rolls to the next game.

Why perfection is so rare here — on purpose

A single game can be a coin flip nudged your way. Twenty in a row, three of them in win-or-go-home January, is a different animal. Because each week multiplies against the last, even an elite roster that projects as a heavy favorite most weeks still runs into the wall of compounding probability — the exact math you can play with on our perfect-season odds calculator. We tuned the engine so that a genuinely great, well-built roster has a real but hard-earned shot at 20-0, a good roster gets there only occasionally, and an average one basically never does. That's the point, and it matches history: since the 1972 Dolphins, no one has finished the job. If everyone went 20-0, the number would mean nothing.

What we deliberately don't do

20-0 is an independent, unofficial project — not affiliated with, endorsed by, or licensed by the NFL, any team, or any other roster-building game that uses a similar name, including any other product called "20-0." The players, teams, and coaches in the game are original, fictional archetypes; we don't use official logos, marks, or uniforms, and we don't claim our sim predicts real games. Historical facts in our guides are for informational and entertainment purposes, drawn from neutral references. It's a film-room thought experiment you can actually play.

Ready to test a roster against all of it — regular season and bracket both? Start your run, or read the strategy guides first.