The math behind every number

1Market value

A player's value is what other managers actually pay for him. Trade Lab blends FantasyCalc (built from real trades) with the expert trade charts, puts every source on the same scale, and weights each one by how fresh it is and how closely its format (1QB or superflex, PPR or not) matches your league. A source that doesn't list a player is left out of his average rather than guessed at. Dynasty leagues add DynastyProcess and rookie draft picks.

Packages are valued the way managers think about them: one star is worth more than two good players that add up to the same number, because the extra player takes a roster spot.

2Projections

Points come from Sleeper's weekly projections, converted to your league's scoring (PPR, half PPR, standard, TE premium and passing touchdown points). Near weeks lean on that week's projection; later weeks lean more on the player's rest-of-season rate. Bye weeks count as zero.

Injuries are handled week by week: a player who is out misses those weeks, and his expected return comes from the injury report, the news and the type of injury, with a small discount for his first game back.

3Lineup math

A trade is judged by your best possible lineup every remaining week, before and after. Points from a player who would sit on your bench don't count, which is why a 2-for-1 can help one team and hurt another. Your bench covers injuries at a typical rate, and an empty spot is filled at the level of players you could find on waivers.

The Lineup view uses the same weekly numbers to set the best lineup for this week, and shows the close calls with a bad-case and good-case range for each player.

4Acceptance odds

"Likely to accept" estimates how often a typical league-mate would say yes. It weighs what they get against what they give on perceived value (managers pay a little extra for star names and for players who were drafted high), whether the deal helps their own lineup, and the cost of taking on extra players. It's a guide to which offers are realistic, not a promise.

5Playoff and title odds

The rest of the season is played out 4,000 times. Each run plays every remaining week with each team's projected lineup that week, random game-to-game swings, and the real schedule when your league shares it (otherwise each team plays the whole field). Teams are seeded by record with points as the tiebreak, then the bracket is played in your playoff weeks. A team's odds are the share of runs it makes the playoffs or wins it all. Records so far count, which is why a strong roster with a bad start can trail a weaker 3–0 team.

The same random numbers are used before and after a trade, so the difference you see comes from the trade alone.

6Buy or sell

Each player's rating runs from strong sell to strong buy and combines up to five signals: his projected production against players priced like him, the latest news and injury report, his 30-day trade value trend (read against the grain when his production doesn't back the move), his recent games against his projection, and how many Sleeper managers are adding or dropping him. A Buy or Sell tag only appears when enough of those signals agree; otherwise it's a lean.

7Dynasty plans

In dynasty leagues every team's strength is projected for this season and the next three, from each player's age curve, and those seasons are simulated (3,000 runs). Your plan, from full rebuild to all-in, comes from your title chances across those seasons and the age of your core. Trades are then judged by how they change title chances in the seasons that matter for your plan, plus long-run value.

8The Trade Analyzer

Without rosters, the analyzer builds typical teams for your settings by drafting the player pool, then places the trade on three of them. In redraft that's a contender, a bubble team and a long shot; in dynasty, a contender, a middle team and a rebuilder. That shows how much a trade depends on the roster it lands on. Everything else (value, projections, odds) uses the same math as the rest of the app.

9Track record

Every finished week, the projections the app used are checked against what happened. A "start/sit call" is two players at the same position whose projections were at least 3 points apart; the call is right when the higher projection scored more. As a yardstick, the same calls are made by points per game so far this season.

Players projected for at least 5 PPR points who played. Scoring is PPR; rankings change little by format.

10Limits

• Everything here is an estimate. Projections miss, values move and managers are people.
• Kickers and defenses aren't modeled.
• Values update as their sources do; during games, values are from before kickoff.
• Trade Lab is an independent tool and isn't affiliated with Sleeper, ESPN, FantasyCalc, DynastyProcess or the NFL.