Pillar — HashmarkIQ
One number. Defensible. Cross-era. Earned.
Madden gives you an overall. PFF gives you a grade. ESPN gives you a Total QBR. Each is useful, each measures a different thing, and none of them know what a third-string undrafted corner who has started three career playoff games is actually worth in this franchise. HashmarkIQ is the rating built for the dataset Hashmark uniquely holds.
What goes in
HashmarkIQ is a position-normalized rating on a 0–100 scale. Every player has one, updated every export. The inputs are the full Hashmark record — not just the rolling Madden snapshot.
- Regular-season production, weighted against position and era baselines
- Postseason production from the Playoff Vault, weighted heavier than regular season
- Draft context: round, age curve, expected ceiling
- Scouting trajectory: how the attribute ratings have moved over time
- Breakout history: number, magnitude, and quality of outlier games
- Durability: snap counts, missed-time patterns
- League context: strength of schedule, scheme fit, opponent quality
What HashmarkIQ is not
It is not a redaction of Madden's OVR with a different brand name. It is not a single-season grade. It is not a future-projection model. It is a rating of who the player has been, across the full longitudinal record Hashmark holds, normalized so that a 78 at left tackle means roughly the same thing as a 78 at slot corner.
A rookie's HashmarkIQ is sparse — the inputs are weighted heavily on the Scout Archive and the synthesized College Years until real production accumulates. A year-12 tight end has a dense HashmarkIQ — 200 career games, 14 playoff games, eight breakout games, and an attribute trajectory we have followed since his combine.
How it's rendered
The rating is presented with its components visible. Hover the number and the breakdown opens — production, postseason, draft context, trajectory, breakouts, durability — each contributing a numeric share. There is no black box. A reader can argue with HashmarkIQ the way a reader can argue with a stat. That is the point.
The methodology document
The full HashmarkIQ specification — weights, position baselines, era normalization, and the changelog of methodology revisions — lives in the documentation set we publish at public beta. The rating is versioned. When the methodology changes, we publish a changelog entry and version the rating column so historical comparisons are valid.
Draft of HASHMARK_IQ_SPEC.md is targeted for Phase 2 (~Week 10 of development). The methodology will be public before the rating ships.