Fantasy Football· Week 8
Trade analyzer
Should you do the deal? Put the players each side gets on their side. Each player's number is the points he's expected to score over the rest of the season beyond what a free pickup at his position would give you. Add up each side and compare. Close is a fair trade. The further apart the numbers, the more one side is giving up, and the better the reason it needs.
Values as of Oct 28, 10:17 AM ET · refresh every morning with the rankings · through week 17
Your league · 12 teams · one QB · full PPR
Roster
Catches
Passing
Rushing and receiving
Every player is re-scored on his own mix of yards, touchdowns, catches and turnovers, and the replacement line moves with his position and your league's shape. Set it once; every tool on the site reads it.
Side A
0.0 pts above replacement- Nobody yet.
Side B
0.0 pts above replacement- Nobody yet.
Add the players each side receives. A side's number is the sum of what its players are worth over the rest of the season, above what a waiver pickup at the same position would give you.
Every player's rest-of-season projection and value
How to read it
The number is rest-of-season points above replacement. I take each player's expected points a game from the same model that builds the weekly rankings, with the week's matchup, spread and weather switched off so what's left is his own volume and efficiency, weighted toward his recent games and pulled toward last season. Multiply by the games his team has left through week 17. Subtract what the last starter at his position would score in those same games, the 14th quarterback, 32nd back, 44th receiver or 16th tight end in a 12-team league. That's his value. A back worth 60 is worth 60 more points than a waiver pickup from here to the end.
Scoring settings. The numbers are built at the usual defaults: 4 a passing touchdown, 1 per 25 passing yards, −2 an interception and a lost fumble, 6 a touchdown, 1 per 10 yards, and receptions by the format you pick. Open "Your league" above the sides, or set it once on the Your league page, and every tool on the site reads it. Each player is then re-scored on his own mix of yards, touchdowns, catches and turnovers, this season's rates pulled toward his position's so a hot month does not set them, and the replacement line moves with the position. A site that scores 6 a passing touchdown by default will show quarterbacks three or four points a game above this one at 4; that is the setting, not a disagreement.
Superflex and two-quarterback leagues. Pick "Superflex / 2 QB" and the replacement quarterback becomes the 26th in a 12-team league instead of the 14th, a man scoring about 14 a game rather than 16 or 17. Every real quarterback's value rises by the difference times his games, which is why nobody in those leagues will trade you one. The other positions do not move; a superflex slot is a quarterback slot in practice.
Why two-for-one works without a thumb on the scale. If you take two players for one, you drop someone to make room, and the player you drop is, by definition, about replacement level, which is worth zero in this math. So the sum on each side already carries the premium a star deserves. I tested curving the values further and it did not help.
How it was tested. For every week 3 through 12 of 2024 and 2025, I valued every player with only what was known that week and checked it against what he actually averaged the rest of the way. It beat a plain season-to-date average at running back, receiver and tight end in both seasons, and beat last year's numbers at every position in both. At quarterback it tied the season average in 2024 and beat it in 2025. Every number is on the model log. Values are graded again on the scorecard when the season ends.
How far off it runs. I measured it. Over those same twenty weeks the per-game number missed what the player actually averaged by about 3 points a game for a back, 2.7 for a receiver, 2.2 for a tight end and 4.2 for a quarterback, in half-PPR. That is the "give or take" under each player: the miss times the games he has left. It's why two backs 40 points apart can read as even. The number also spreads players out more than the season does, by about a fifth at back and receiver and more at quarterback, so the gap between two stars is usually smaller in October than the tool makes it look. I tested leaning harder on last season to fix that and it helped one year and hurt the other, so the number stays and the error is shown instead.
Who plays. An injured player's weeks are scaled by his chance of playing them, measured from five seasons of injury reports: Out and Doubtful are nothing this week (0 of 1,426 and 1 of 231 played), a Questionable starter plays 70% of the time, injured reserve is nothing until the league's four games have passed and then the return rate for what is hurt, and a torn ACL or Achilles is the season. The line under each player says what was done with him.
What it can't know. The actual day a player returns; the odds are averages over hundreds of injuries. Which weeks your playoffs are. Whether the man you're trading for is about to lose his job. Read the line under each player to see where the number came from. Nothing here is advice; it's the arithmetic, shown.