Our model’s read of every player, blended 60/40 with the market. Separate boards for 1QB, superflex, half-PPR, TE-premium and 3-WR leagues. When we disagree with the market, history says we’re right 63% of the time. How we test it.
Static values
Every number below is our board, for your format. Each player is valued on each team’s own timeline, because a contender and a rebuilder are buying different things. Your format (1QB or superflex, PPR, TE premium, 3 WR, passing-TD points) sets the board; team count and starting slots move only contending and rebuilding values.
The Read
Add players to both sides
Favors Team AEvenFavors Team B
Team A — balanced
—
Add players to both sides.
Team B — balanced
—
Set each team's window above — the same trade can be right for both sides.
Opens for anyone, and values the trade in their own settings.
Rookie picks, priced three ways. The same pick is a different asset to a team trying to win now and a team playing for 2029. The prices are the same in every format; the player each pick trades for comes from the board for yours.
The board is rebuilt every Tuesday. Everything below is measured against
the last one that was published — not against a hand-written list, and not against last year.
Two managers in the same window can only take from each other. A deal that
helps both sides needs two different timelines — so that is what this searches for, and it skips
the managers who want exactly what you want.
How the board is built. Our board blends two views. Our model (60%) reads each player’s usage, production, age, draft capital, team situation and where dynasty drafters are taking him, trained on 13 seasons of outcomes. The market (40%) is what players fetch in real trades, blended with FantasyPros’ dynasty expert rankings; in season the trades count three times as much (trade values from FantasyCalc, RosterAudit and, in superflex, Dynasty Dealer), refreshed every Tuesday and Friday; before the season, dynasty draft data counts too. Tap or click any player for the full case. How it works →
How the board is tested
We rebuild past seasons’ player tables exactly as they stood each August — no later stats, rosters or contracts — build the board from them the same way, and check it against what those players went on to score in the seasons that followed.
Colour keyQBRBWRTE
Import it, search it, or paste it. Sleeper leagues import in one click, settings and all. Everyone else can search players by name or paste a roster — anything that doesn’t match gets a one-click fix. Nothing is stored; the whole read happens in your browser.
What To Do About It
In order. The first line is the one that costs you the most.
Your Starting Lineup
Under your league’s slots. The grade is this season only — where he ranks at his position against how many start in your league
On The Bench
Value that never takes the field — this is what you trade with
Positional Strength
Long-term value, slot by slot, against what an average team in your league starts there. A veteran can grade well above in your lineup and below here — that gap is the whole reason to trade him
The Clock
What this roster is worth in three years if you stand pat
Sell Candidates
Worth more to somebody else than to you — move them while that is true
Buy Targets
Cheap for a team on your timeline, expensive for the market
Your read appears here
Put a roster in — a Sleeper username brings the roster and your league’s format across in one click — and the model answers, in order:
Which window you are inContend, hold, or rebuild — read from the roster, not from your record
What to do about itThe moves, ranked by what each one is costing you
Lineup & benchEvery starter graded for this season; every bench asset priced as trade capital
The clockWhat this roster is worth in three years if you stand pat
Sell candidatesPlayers worth more to somebody else’s timeline than to yours
Buy targetsCheap for a team on your timeline, expensive for the market
Nothing is stored. The whole read happens in your browser.
Import your league, or paste it. A Sleeper username pulls every roster and your real format in one click. To paste instead: one block per team, the manager name on the first line, then the roster, with a blank line between teams. The first block is read as your team — or click any row on the board to switch. Nothing is saved unless you sign in.
Sample league. Twelve anonymised rosters, so you can see what the read looks like. Import or paste your own league to see yours.
The Board
Every roster, ranked by win-now strength. Click a team to make it yours — everything below re-reads from that team's side.
Win-now — best starting eight at a contender's discountFuture — top fifteen at a rebuilder'sBars run from this league's weakest roster to its strongest. The number is that team against the league average, where 100 is average.
TeamWindowWin-nowFutureAge
Roster
The selected team, lineup then bench, under this league's slots. A grade is this season against what actually starts at that position in this league — a B on the bench is a starter somebody else needs. Click a team above to switch.
Where You Stand
Where your starters rank against the other eleven teams at that spot — a full bar is first in the league, an empty one is last
Areas To Improve
Read in order — the first line is the one that costs you games
Calls To Make
Trades where both managers gain against their own window, within 18% of each other on our balanced board so neither side is being fleeced
The Rest Of The League
Deals that do not involve you — who else is about to get better, and which of your rivals is the one making the call
Your league appears here
Import it — a Sleeper username pulls every roster and your real format in one click — or load the sample league to see what the read looks like first. You get:
The boardEvery roster ranked on win-now strength and on future value, with the window each team is actually in
Where you standYour starters against every other team, position by position
Areas to improveRanked by what each one is costing you in games
Calls to makeTrades with a specific manager where both sides gain against their own window
The rest of the leagueWho else is about to get better, and which rival is the one making the call
Nothing is saved unless you sign in.
How it reads. Every player is valued off our board: our model’s read of his usage, production, age, draft capital, team situation and where dynasty drafters are taking him, blended 60/40 with the market (real-trade values and expert rankings, plus dynasty draft data before the season), on the board for the format you set above — superflex has a board of its own, which is why a quarterback is worth more there than in 1QB. Contender versus rebuilder is how much the future counts: a contender values a season three years out at about a quarter of this one, a rebuilder at about one and a half times this one. Each side of a trade is scored on its own timeline, which is why both columns can read positive at once. Rookie picks are valued as the player their historical slot returns, hedged and discounted for the wait (a rebuilder waits for free). Trade verdicts are the percentage each side gains against what it sends away, judged on its own timeline.
Override Board
The engine owns the order. Drag a player, or set his override directly, when you know something the box score does not. Every change is logged and goes live for everyone the moment you save.
Drag a player by the handle to move him up or down.
+0.0
Engine points, added on top. This survives every weekly recompute and every slider — it is the one place a human overrules the algorithm, so it goes in the log.
Bruiser, balanced or receiving. Feeds the career-wear curve only.
