The source stack is deliberately small, attributable, and cheap to serve. New feeds only land when their rights, identifiers, and update paths are clear.
Composite market values, ranks, attributed history when supplied, and raw season totals used by the transparent scoring adjustment. A scheduled publisher validates the licensed feed and packages the current release with the app.
Read official API docs ↗All 32 teams, the complete current schedule, venues, rest, surfaces, and prior-season scoring results. FTT records source hashes and derives the visible matchup-temperature model under CC BY 4.0 attribution.
Inspect the data project ↗Scoring-relative player values, roster-shape and FLEX replacement, head-to-head comparisons, raw totals, roster-cost weights, verdict thresholds, share URLs, and balancing suggestions execute locally and deterministically. They do not require a model call or third-party trade-evaluation request.
Inspect the formula →Player pages use Wikimedia API-selected images with reviewed Creative Commons terms. When no clearly reusable portrait exists, the profile uses original FTT player-file artwork instead. Every record retains its creator, source, and terms; Sleeper and ESPN player-image hotlinks are not used.
Visit Commons ↗Player pages combine current rosters and injury reports with regular-season weekly statistics and offensive snap counts from nflverse. The app currently covers 220 roster matches and 210 player game histories across 2024, 2025, 2026.
Fantasy Trade Target reads the project's versioned release files directly, records a hash and row count for every input, and packages only the fields used on the site. Season summaries retain the complete three-season window; player tables show the ten most recent games. The release is refreshed with the rest of the scheduled data publication.
The underlying data is attributed to nflverse under CC BY 4.0. The current integration does not use player headshots or FTN charting data, and it does not turn an injury or practice row into a medical prediction.
The player-comparison collection is a new view over the same current release, not a second ranking feed. Each page joins two reviewed player profiles with the current dynasty Superflex, dynasty 1QB, Superflex TE-premium, and redraft markets, then runs the published scoring engine for the relevant setting.
Visible production fields come from the licensed profile record. Comparison conclusions are deterministic value-gap statements generated from the displayed numbers. The hand-reviewed matchup manifest contributes only the pairing, decision frame, and explanatory context; it never overwrites market or scoring values.
The rookie-pick collection reads the 2027 first and second rounds directly from Tradyr's separate 1QB and Superflex pick markets for 8-, 10-, 12-, 14-, and 16-team leagues. FTT never estimates a missing slot or converts one league format with a generic multiplier.
Every successful full-market archive also contributes a compact public observation for each covered pick: UTC time, release ID, 12-team 1QB value, and 12-team Superflex value. The initial 28-release backfill begins August 4, 2026, and the scheduled publisher appends new validated observations without rewriting older release IDs.
Player equivalents, adjacent-pick gaps, and cross-year matches are deterministic nearest-neighbor views over that same release. They add decision context but never overwrite the source value. Each exact-pick page exposes both a machine-readable JSON record and complete history CSV.
Tradyr’s composite market is always the anchor. FTT combines that market with licensed raw season totals, calculates fantasy points under the selected rules, and adjusts only for the player’s change relative to replacement at the same position. No generated opinion or hidden projection changes the score.
Passing, rushing, receiving, turnover, and two-point-conversion totals from up to three recent seasons become confidence-weighted per-game profiles.
PPR means points per reception. Passing TD: 4 or 6 points. Standard: 0 per catch. Half PPR: 0.5 per catch. Full PPR: 1 per catch. Yardage and touchdown scoring otherwise stays fixed.
The model compares each player with replacement after dedicated starters and FLEX demand are allocated, scales for sample confidence, and caps movement at ±12% in dynasty or ±20% in redraft. Picks and players without usable profiles remain unchanged.
Player values and ranks are captured throughout the day, versioned, and retained for market movement, downloadable records, and long-term research.
The latest validated market variants, compact scoring profiles, reviewed player profiles, and profile histories live in the versioned public release used by every server-rendered page and calculator.
Timestamped player value and rank records plus separate 1QB and Superflex observations for 24 exact rookie picks are retained chronologically and exposed through public downloads.
Complete market and pick variants are stored as compressed, append-only snapshots in the private build repository. Raw archives are not shipped in the production container and can move to immutable object storage without changing public URLs.
Scoring, roster slots, taxi, IR, picks, standings, and manager-specific needs.
FTT-owned comparisons between market movement, opportunity, usage, and production.
An anonymized, rights-reviewed corpus of completed prices by format and league size.