Fantasy football has a schedule problem. You play one opponent a week, drawn at random, so a good team that keeps running into the week's high scorer looks mediocre, and a mediocre team that keeps drawing the week's worst looks like a contender. All-play removes the schedule.
How it is calculated
Every week, compare your score against all the other teams that week, not just your opponent. In a 12-team league you get 11 results per week instead of one.
Week 4, you scored 118.4 beat 9 teams lost to 2 teams real record that week: 1-0 (you played the 130-point team? then 0-1) all-play record that week: 9-2
Do that for every week and you have a record built on how you actually scored, with the schedule taken out of it.
Luck: the gap between the two
Subtract your all-play win percentage from your real win percentage. A positive number means the schedule has been kind; a negative number means you have been scoring like a better team than your record shows.
real win% = 0.643 (9-5) all-play win% = 0.519 luck = +12.4 percentage points
That manager is 9-5 while playing like a 7-7 team. It is not an accusation of anything — nobody chooses their schedule — but it does predict a correction.
How to read it
- Large positive luck with a good record: the record is flattering. Expect regression.
- Large negative luck with a bad record: the team is better than it looks. Buy low in trades.
- All-play near .500 across many seasons: genuinely average, whatever the trophies say.
- One season is noise. All-play only becomes convincing across several years.
All-play needs week-level scores, so it can only be computed for seasons your platform actually recorded. Hand-entered seasons that kept only standings cannot produce it.