Fantasy Football Power Rankings Generator — Free, Any League

How it works

You type your standings in — team name, wins, losses, ties, points for, points against — and the math runs in your browser. That's the whole pipeline. Nothing connects to your league, nothing is uploaded, and no account is involved on any platform; the table is saved on your own device so next week you're only updating five numbers a team. Every fantasy host prints exactly these columns, so the same page works whether your league lives on Sleeper, ESPN, Yahoo or NFL.com. There's a paste box too — select your standings table, copy, drop it in, and it will pull the rows out.

Why all-play beats your W-L record

Your record measures two things at once: how good your team is, and who you happened to play. A 100-point week is a win against the team that scored 90 and a loss against the team that scored 130 — same team, same lineup, different record. The all-play record deletes the schedule from the equation: instead of the one opponent the schedule handed you, your team is measured against every team in the league. In a 12-team league that's 11 results a week instead of one, which is why all-play converges on true team strength far faster than the standings do. It's the difference between "who did you beat?" and "how many teams would you have beaten?"

From season totals, all-play is computed at your scoring average: your points per game against every other team's points per game, counted once for each game you've played. That's exactly what the week-by-week version collapses to if every team scored its own average every week. It costs you the week-to-week variance — a boom-bust team and a metronome with the same average look identical here — and in exchange it takes five numbers instead of a full season of box scores.

That makes it an estimate, and the tool labels it one — in the rankings and in the group-chat copy. Because it is built from points per game, an estimated all-play is a straight line in scoring order: in a 12-team league the top scorer reads 33-0 and the last-place scorer 0-33, whatever either did week to week. If you want the real count, the optional weekly-scores box takes each team’s score for every week played and counts all-play week by week instead, with your actual last three weeks as the recent-form term.

The luck index

Once you have all-play, luck falls out of a subtraction:

luck = actual wins − (all-play win% × games played)

The second term is how many games you should have won given your scoring — your expected wins. A team that's 6-3 with a 55% all-play through nine weeks was expected to win 0.55 × 9 = 4.95 games, so its luck index is 6 − 4.95 = +1.05: about one full win gifted by the schedule. Positive is lucky, negative is unlucky, and the most negative team in the league gets the CERTIFIED UNLUCKY badge to screenshot for the appeal. Ties count as half a win on both sides of the ledger.

Luck, week by week

With weekly scores the season luck number can be taken apart by week. Each week a team should win as often as the share of the league it outscored: beat 7 of the 9 other teams and the week was worth 0.78 of a win. The week's luck is what happened minus that. A team that posts the second-highest score in a 10-team league and loses has a week of about −0.9; a team that posts a bottom-three score and wins has a week that lucky in the other direction. Add the weeks up and you get the season luck index above. To see the weeks, put a W, L or T after each score in the weekly box (Purple Reign 121.4 W 98.2 L). A grid of scores does not say who played whom, so the result has to come from you. The group-chat copy then adds the unluckiest and luckiest week in the league.

The power score, disclosed in full

There's no secret sauce — the ranking is exactly this:

power score = 0.5 × norm(PPG) + 0.3 × all-play win% + 0.2 × norm(last-3-weeks PPG)

• 50% points per game — scoring is the most predictive single stat in fantasy; teams that score keep scoring.
• 30% all-play win% — the schedule-proof record, so consistent scorers beat boom-bust teams with the same average.
• 20% recent form — normally last-three-weeks PPG, so a September roster that got hot in November isn't ranked on ancient history.

norm() is min-max normalization within your league: the best mark gets 1, the worst gets 0, everyone else lands proportionally between. That keeps the ingredients on the same scale before the weights are applied — and it means the power score only ever compares you to your own league, not to some global average.

The one thing season totals can't give you is that third term. A standings page has no week-by-week detail, so recent form falls back to season pace and that 0.2 rides along with the 0.5 (paste weekly scores and it becomes your real last three weeks). Said plainly: the board you get from season totals alone is 70% scoring, 30% all-play, which is a scoring-order ranking — and the luck index is where it stops agreeing with your record. That disagreement is the whole product. The arrow beside each team is not week-over-week movement; it's this ranking against your own standings, so ▲3 means the power board has you three spots higher than your record does.

The fine print

• Median-scoring leagues (a second weekly game against the league median): tick the median box and type your records the way your platform shows them, both games counted. Points per game then divide by the weeks played, not the results. The luck index counts the head-to-head game only, because the median game is your own score against the whole league and carries no schedule luck. Splitting the two needs the weekly scores; without them a median league gets rankings and all-play but no luck number. The group-chat copy says it is a median league.
• Type finished weeks only — enter the standings as your platform shows them once the week is scored. A half-played slate on the standings page (say, Friday after Thursday Night Football) will read as real points here, because the page can only believe the numbers you give it.
• The numbers get sanity-checked, not corrected — if league-wide wins don't match losses, or points for don't match points against, you get a note pointing at the typo. Nothing is silently "fixed": the rankings always use exactly what you typed.
• Before Week 1 there's nothing to rank — with no games played there's no scoring to rank on. Come back once real points are on the board.

Sorting out dues instead? The league payout calculator splits the pot, and the FAAB bid calculator prices the waiver claims that change next week's rankings.

FAQ

Where do I find these numbers?

Your league's standings page has all five columns already: team, record, PF, PA. On Sleeper it's League → Standings; on ESPN, Yahoo and NFL.com it's the league home page. Select the table, copy it, and drop it into the paste box — or just type the rows, which takes about a minute for a 12-team league and is then saved on your device.

Does this connect to my league?

No, and that's deliberate. It never asks for a login, a league ID, or a permission of any kind, and it can't see or change anything in your league. Every number on this page came from your own keyboard, the math runs in your browser, and the only thing stored is your table — in your browser's local storage, on your device, so you're not retyping it next week. Clearing it is one tap on Reset.

Why don't these match my standings?

The standings sort by W-L, which bakes in schedule luck. This ranking sorts by scoring — that's the point, and the arrow beside each team measures exactly how far apart the two answers are. If your league plays a weekly game against the median, your record already includes those results: tick the median box and the tool accounts for it.