GIQ Methodology

One player model, translated into the settings that change fantasy value.

GIQ separates the football projection from the league context around it. That lets the same underlying player evaluation behave differently in Redraft, Dynasty, Keeper, PPR, Standard, 4-point passing-TD and deeper-league environments without copying market ranks.

Methodology for the 2026 Week 1 release

1. Start with the football projection

GIQ estimates expected fantasy production from a player's workload, efficiency, scoring opportunity, receiving or rushing contribution, team context and expected games. The model also keeps a likely floor and ceiling around the central projection so two similarly projected players can still have very different risk profiles.

2. Evaluate the profile, not only the point total

The ranking system uses a broader set of signals than projected points alone. The current framework includes forward projection, sustainable projection, production/output, role/volume, role stability/security, efficiency, passing quality, rushing value, team environment, health/availability, age/timeline and draft-capital insulation. Not every category matters equally to every position or format.

3. Translate value by league type

Redraft is dominated by the coming season. Dynasty gives more weight to multi-year durability, age curves and long-term role insulation while reducing the tendency for one season of positional scarcity to overwhelm asset value. Keeper is intentionally closer to Redraft, with a smaller future-value component because generic keeper rankings do not know each manager's specific keeper cost.

4. Translate value by scoring

PPR, Half PPR and Standard are not just three labels on one board. Reception scoring changes the value of catch volume. Four- versus six-point passing touchdowns changes quarterback scoring directly. League size changes replacement-level scarcity, which can alter how much value a scarce position adds relative to an alternative at another position.

5. Market is a QA alarm, not a score

GIQ can compare its output to external rankings after the board is built. A large disagreement is a reason to investigate the player's role, projection, availability and model treatment. It is not a reason to move the player automatically. This keeps the board independent while still using outside disagreement as a way to find possible mistakes.

6. Personalization stays separate from True GIQ

Quick Pick learns how a user breaks close ties. After at least 20 decisions it can build a personal board with movement guardrails. That personal board never rewrites True GIQ. The distinction matters: the model's opinion and the user's preference are both useful, but they are not the same thing.

Current public scope: 504 certified public player assets, 72 boards and 35,952 certified ranking rows across Redraft, Dynasty and Keeper.

What GIQ does not claim

GIQ does not claim that a ranking is certain, that a floor or ceiling is a confidence interval, or that one model can eliminate football uncertainty. The purpose is to make the assumptions visible and consistent enough that a user can understand why the board changes when the league changes.

Read the full research note on how GIQ rankings work →