Every Resolve NFL advanced metric, with what it measures and how to read it. Each metric had to pass a face-validity gate before shipping — if a metric ranked someone obviously wrong, it got cut. Grouped by side (offense / defense / two-way / team).
Offense
Pressure-Adjusted CPOE (DIME)(DIME)
Strip the offensive line out of QB accuracy.
What it measures: Completion % Over Expected with each QB projected to the league-average pressure rate. Pressure rate is mostly OL quality — removing it isolates the QB's accuracy from the pocket they're playing behind.
How to read it: Russell Wilson #1 career since 2018 (3,121 attempts, +0.047). Mahomes #9 career — his peak is top-5 but his floor seasons (2021, 2024) drag the average. Brock Purdy #19 career answers "is Purdy real?" with a number: mid-tier real, not the offensive system carrying him.
per-QBlive
Earned-Separation YPRR (ISO)(ISO)
Beating coverage vs getting schemed open.
What it measures: Yards Per Route Run on plays defended in man coverage. Isolates a receiver beating a defender one-on-one from the yards their scheme gave them.
How to read it: Cooper Kupp 2021 = 12.40 vs man (3rd-highest in the dataset) — his triple-crown season was earned, not schemed. Puka Nacua 2023 = 10.65 (rank 7) — locks his breakout as real. Career top tier: Chase, Hill, Jefferson, Julio Jones.
per-receiver (WR/TE)live
Schemed YPRR (FREE)(FREE)
The other side of the ISO coin.
What it measures: Yards Per Route Run on plays defended in zone coverage — the "scheme handed me this look" counterpart to ISO. The pair together decomposes a receiver into earned-vs-schemed production.
How to read it: High FREE with lower ISO = scheme-dependent archetype (Amon-Ra St. Brown career earned/schemed ratio 0.909 — better vs zone than man). Ja'Marr Chase 2024 ratio dropped to 0.707 — coverages adjusted; he became zone-dependent that year, a real arc the metric catches.
per-receiver (WR/TE)live
Pressure Suppression vs Expected (TTPR)(TTPR)
Mobile QBs hold the ball forever and STILL escape pressure.
What it measures: Pressure rate minus the rate you'd expect from time-to-throw. Mobile QBs hold the ball long but evade — they suppress pressure relative to expectation. Pocket QBs whose OL gives up extra pressure show up high.
How to read it: Lowest TTPR (most pressure suppressed): Kyler Murray #1 (-0.075), Rodgers, Lamar, Caleb Williams, Hurts — the evade-style QBs. Highest TTPR (excess pressure beyond what TTT predicts): Daniel Jones, Bortles, Cousins, Tannehill, Matt Ryan, Rivers — pocket QBs whose lines surrendered more pressure than time-held implied.
per-QBlive
Fumble Luck (FRLI)(FRLI)
Who's been lucky on the bounce, and who's owed a few back.
What it measures: Lost-fumble rate vs the league baseline of 44.4%. Negative = lucky variance (lost fewer than expected, due for regression). Positive = unlucky variance (lost more, due for regression up).
How to read it: Big regression-down candidates: Kyler Murray (50 fumbles, 24% lost — half league rate), Justin Fields (44 fumbles, 27% lost), Jordan Love. Big regression-up: Travis Kelce (70% lost on 10), Dalvin Cook (71% on 21 — real sample). This is fantasy gold for RB/QB redraft.
per-fumblerlive
3rd Down Conversion Above Expected (STIX)(STIX)
Who actually moves the chains on third down.
What it measures: Conversion rate on third down minus the expected rate given distance and game state. Strips out garbage-time short-yardage and isolates real third-down ability.
How to read it: Mahomes #2 (+0.080 over 1,202 attempts), Jimmy Garoppolo #3 (+0.075), Brock Purdy #4 (+0.061) — the SF system QBs cluster (real or scheme). Tua #6 (+0.046 on 511 attempts) supports the "real regular-season QB" read. Bottom: rookies + bad single-seasons (Bryce Young, Z. Wilson, Howell, Levis).
per-QBlive
YAC Above Expected (RAC)(RAC)
Who turns 5-yard catches into 25-yard plays.
What it measures: Yards-after-catch minus the expected YAC for each reception (based on catch location, defenders nearby, etc.). Isolates the receiver's separation + run-after-catch from the throw they got.
How to read it: George Kittle #5 (480 rec, +2.30 RAC — elite TE), Austin Ekeler #14 (+1.56), Puka Nacua #18 (+1.19), Ja'Marr Chase #21 (+1.09), AJ Brown #25 (+1.05), Gronkowski #22 (+1.07). Bottom: late-career Larry Fitzgerald (-1.48), Witten, Crabtree — contested-deep specialists who give back YAC.
per-receiver (WR/TE/RB)live
True YPC(True YPC v2)
Strip the box-defender count out of YPC.
What it measures: Box-Adjusted YPC — projects each RB's carries to the league-average distribution of box defenders, so a back who faces stacked boxes doesn't get docked for it.
How to read it: Derrick Henry career +0.174 box adjustment — he faced 0.4 more defenders than league average, raw YPC under-rated him by 0.17. True YPC: 5.03. Top-15 careers: McCaffrey, Henry, Chubb, A. Jones. Saquon Barkley sits OUTSIDE top-15 — his Eagles breakout doesn't carry his career after the box adjustment.
per-RBlive
Defense
Defender Coverage Snaps Allowed (DCSA)(DCSA)
IDP coverage isolation — who actually shuts WRs down.
What it measures: Yards per coverage snap a defender allowed across his career. The defensive-back equivalent of an opponent quarterback rating — isolates coverage skill from box-stat noise.
How to read it: Patrick Chung #2, Hightower #3, Brandon Carr #4, J.C. Jackson #9 (1,895 plays), Chauncey Gardner-Johnson #13 (2,270 plays — best big-sample tier), Quinyon Mitchell #20 (rookie hit). Bottom: Ivan Pace, Zaven Collins (1,145 plays — sample-size legit), Nick Cross. Most public DB metrics don't isolate coverage snap%.
per-defender (IDP)live
Team Coach
Coaching Tree(Coaching Tree)
Mentor-mentee edges — who produced whom.
What it measures: Career mentor-mentee count per coach. Built from 3,200 lineage edges across the modern NFL coaching record. Informational sidebar — useful as context when reading HC-hire predictions.
How to read it: Top mentor-trees by mentee count: Mike Nolan (16), Gregg Williams (15), Schottenheimer/Zimmer/Olson (14). McVay tree: 7 mentees (Barry, Fassel, Lake, Phillips, Callahan). Reid tree: 6. Belichick: 5 (Flores, Judge, Patricia, Crennel, Groh). Sean Payton: 5. Use as background context, not a predictive signal.
per-coachlive