Skip to content

Trade Finder

The Trades tab on the My Team page answers the question the rest of the app doesn’t: “Who do I even approach, and with what?” It scans every roster in your league and proposes specific, mutually-beneficial trades — pairing your category surplus against a rival’s deficit, and vice-versa.

Every proposal is one click from the AI Trade Verdict, so you can grade and share it before you ever message another manager.

The Trade Finder builds each proposal from three tests. Only offers that pass all three surface:

  1. Category complementarity. It compares your projected category totals against the league average to find where you’re strong and where you’re weak — then looks for a partner with the mirror-image profile. You need what they have a surplus of; they need what you can spare.

  2. Concrete players. From the partner’s surplus it picks a real, tradeable player who embodies the category you need (and symmetrically picks the player you can afford to send). It builds 1-for-1 packages first, then 2-for-1 and 2-for-2 to balance value.

  3. A fairness filter. A package is only shown when the two sides’ value is within a tolerance band. Lopsided “steals” are filtered out — these are offers a real manager might actually accept. Player value here is current-form-aware: a consensus ranking eased toward the player’s actual current-season production (weighted by league format and season progress), so a slumping star is valued for how he’s playing now, not his preseason billing. Players without enough of a sample fall back to pure consensus.

Crucially, both teams’ category outlook must improve for a proposal to appear. Nothing one-sided is ever surfaced.

The Trade Finder also respects injuries: it will never propose that you receive a player who is on the 60-day IL or otherwise out for the season, so you’re never offered an unavailable player.

Each card shows:

  • You send → You get — the concrete players on each side, with consensus ranks.
  • Category gains — which categories you improve, and which the partner improves.
  • A value-balance chip — “Even value” when the two sides line up, or who comes out slightly ahead.
  • Rationale — a one-sentence pitch for why the deal works. When it carries a small ✨ sparkle icon, the sentence was written by AI with the full context of both rosters; without the icon it’s a templated summary of the category math. (AI rationales require the server to have an Anthropic API key configured — without one, the templated version appears and nothing else changes.)
  • Grade this trade — opens the AI Trade Verdict flow pre-filled with both player sets. Generate a verdict, then share it.

Only fantasy team names are ever shown — never manager real names.

The Trade Finder is format-aware:

  • Redraft leagues match purely on category needs and value. Player age and long-term outlook are ignored — the only question is who gets more rest-of-season production.
  • Dynasty and keeper leagues add a competitive-window term on top of category matching. Proposals that align each team with its window — a contender receiving win-now production while shipping youth to a rebuilder — are rewarded and flagged with a Window-matched badge. This uses the same window scores as the State of the League index.
  • Proposals are computed on demand from your league’s current rosters and projections. If rosters or projections have just changed, give the data a moment to refresh.
  • The search is bounded to the best-fitting trade partners for speed; when it’s bounded, the tab notes how many partners were searched.
  • No trades found? That’s normal — it means there’s no roughly-fair, mutually-beneficial deal available right now, not that no trade is possible. Check back after the next roster or projection update.