It's the Tuesday after Week 1. Somebody in your league put up 188 points and has spent the last 36 hours trying to figure out how to brag without risking jinxing his team. Somebody else scored 71, and the group chat has already held his team's funeral.
Everyone is asking the same question, and it has an actual answer: how much does one week tell you?
I measured it on 846 real, finished leagues — 9,744 team-seasons from 2018 to 2025. For each team I took its week-1 score, measured it against its own league's average that week, and then looked only at what happened after. Every percentage below is for a twelve-team league with six playoff places.
The short answer: about a tenth of it is real
Every point you score above your league's average in week 1 predicts about a tenth of a point a week for the rest of the season. Put a full 20 points (1 standard deviation) on them — a clearly good Sunday — and it predicts 1.8 points a week above average across the thirteen weeks that follow.
A completely separate measurement lands in the same place. Across those leagues the durable difference between a good fantasy team and a bad one is about 7.5 points per week. The week-to-week noise inside a single team — the same roster, bouncing around its own average — is 21.4 points. That makes one week about 11% signal.
One week of fantasy football is roughly one part information to nine parts noise.
So why did your playoff odds really move?
Because week 1 does two entirely different things at once, and people conflate them.
It reveals a little about how good you are — the tenth above. And it banks something regardless: points in the column, and possibly a win. The banked part is not a prediction at all. It counts in the standings whether or not it ever repeats.
In points: a 20-point week 1 banks you 20 points and forecasts about 23 more points during the regular season. Split by their effect on playoff odds, 56% is the head start you banked and 44% is what the score revealed about the team.
The thing I didn't expect: the returns are lopsided
Here is every team by how far its week-1 score sat from its league's average, and what followed.
| Week 1 vs league average | Make playoffs | Points per week afterwards |
|---|---|---|
| More than 40 below | 30.0% | -7.0 |
| 20 to 40 below | 33.7% | -2.5 |
| 0 to 20 below | 44.3% | -0.6 |
| 0 to 20 above | 57.0% | +0.9 |
| 20 to 40 above | 64.3% | +2.3 |
| More than 40 above | 64.7% | +1.4 |
Look at the top row. Then look at the bottom row. They are not even close to mirror images.
Above average, the returns stop. A team 20 to 40 points above its league makes the playoffs 64.3% of the time. A team that goes completely nuclear — more than 40 points clear of everyone — makes them 64.7%, and actually scores slightly less afterwards. Past a certain point, a huge week 1 stops telling you anything new. You have already learned everything a good week has to teach by the time you are 20 points clear; the rest is a random receiver scoring two long touchdowns.
Below average, it keeps going. A team 20 to 40 points short gives back 2.5 points a week for the rest of the season. A team more than 40 short gives back 7.0 — nearly three times as much, off a similar-sized miss.
I did not expect this, so I went looking for the mechanism somewhere I control everything: my own simulator, which plays real NFL weekly scores through drafted rosters across sixteen seasons.
In a room where every manager is equally diligent, the effect vanishes. Over 1.15 million simulated team-seasons the best week-1 scorer finishes +0.28 and the worst -0.28, in standard deviations — symmetric to three decimal places. Week 1 is noise, and noise is even.
Then I ran it again on a room with a deliberate spread of manager engagement, where a quarter of the seats draft loosely and never touch the waiver wire. The asymmetry appears, and the internals say exactly where it comes from:
The disengaged managers are 25% of the room. They are 33.9% of the teams that score worst in week 1, and only 18.7% of the teams that score best.
And when one of them does score worst in week 1, its rest-of-season form is -0.69, against -0.21 for an attentive manager who had the same bad Sunday.
That is the mechanism. A catastrophic week 1 is not a random draw. It over-samples the teams with something durably wrong with them. Those conditions persist. A monster week 1 over-samples nothing: there is no equivalent way to be accidentally excellent for four months.
The obvious objection: a team that starts badly loses interest, stops setting a lineup in November, and drags its own season average down. That would produce this pattern without any of it being knowable in week 1.
It is not what happened. I split the season and looked at weeks 2 to 6 only — before anyone is mathematically buried, while everybody is still paying attention. The asymmetry is strongest there, not weakest: over those five weeks the worst week-1 team runs at -0.24 standard deviations while the best week-1 team runs at +0.03.
Sit with that second number. Over the following five weeks, the team that scored more than anyone else in the league on opening weekend is, on average, an exactly average team.
When should you actually believe what's happening?
Evidence does accumulate, just slowly. This is the correlation between what you have scored so far and what you go on to score:
| Evidence so far | Correlation with the rest of the season |
|---|---|
| After 1 week | 0.14 |
| After 2 weeks | 0.24 |
| After 3 weeks | 0.30 |
| After 4 weeks | 0.32 |
| After 5 weeks | 0.36 |
| After 6 weeks | 0.38 |
Week 2 is worth nearly as much again as week 1, and by week 3 you have roughly doubled what you know now. Then it flattens hard: weeks 4, 5 and 6 add about as much between them as week 2 added alone. Even six weeks in, scoring explains well under half of what your team goes on to do.
What to do on Tuesday
Cross your week-1 result with your week-1 score and you get the only four-box grid that matters:
| Your week 1 | Make playoffs |
|---|---|
| Won, scored above average | 62.5% |
| Won, scored below average | 54.3% |
| Lost, scored above average | 49.1% |
| Lost, scored below average | 36.6% |
Lost but scored well? You are at 49.1% — a coin flip, which is exactly where you started the season. You drew the league's high scorer. Do nothing. It is the most over-reacted-to box on the grid, and 11% of all teams land in it.
Won but scored badly? 54.3%, and your rest-of-season form is dead average. You are not good; you are ahead. Those are different, and the standings will not tell them apart for another month.
Lost and scored badly? 36.6%. Not fatal — more than a third of these teams still make the playoffs — but this is the one box where the number has moved enough to be worth your time this week. And the further below average you were, the more likely it is that the problem is something you can actually name: a starter who is hurt, a position you punted, a first half injury to Sam Darnold and AJ Brown. Go and find it. The teams in this box that turn out fine are the ones where the answer is "we ran into a buzzsaw." The ones that do not are the ones where nobody went looking.