Week 1, 1991. Randall Cunningham's knee goes out on the first play of the second quarter, taking my entire year with it. I didn't lose because I scouted performance wrong. I lost because nobody was even asking the injury question.
Thirty-five years later, here is the tool I wanted that morning. Every NFL player, scored 1 to 100 "Frailty Score" for how likely they are to break down this season. Ironman at the bottom. Glass at the top. Type a name and find out.
No account. No card.
Try 5 free picks — search any NFL player, get a real frailty report. No account, no card.
This is what you get after signing up. Live data, real players, updated continuously.
Position-aware frailty scores for every rostered player — offense, OL, and defense. Sort, search, and compare across the entire league.
In-depth breakdowns with injury history, workload analysis, position context, and actionable durability forecasts.
Historical data from 2000–2026 informing the model. Weekly stat refreshes keep your edge sharp all season long.
By the time ESPN flashes that injury update, the prop has closed, the DFS lineup is locked, and your trade window is gone. You're reacting to news alongside the crowd instead of managing risk before it materializes.
Use your own rankings, your own picks. INJSUR sharpens your decisions — it doesn't replace your judgment.
DFS cash games, prop betting, draft day, dynasty, futures, auction — same frailty score, different edge depending on your game.
Explore all 7 use cases →Our proprietary algorithm scores every skill-position player on a 0–100 frailty scale. Higher score = higher injury risk. Built from workload, contact rate, age, injury history, and positional hazard data — not expert opinions or medical reports.
A logistic-regression and neural-network ensemble informed by 25 seasons of NFL data (2000–2026), weighting each player's notable-injury records from the last five seasons (2022–2026) most heavily. It scores every player based on our trained model output across 20 features: position-specific hazard rates, cumulative workload (carries, targets, sacks, tackles), contact rate, negative-play frequency, YAC exposure, age, experience, and five recency-weighted injury-history variables. This creates a robust, accurate, world-class predictive algorithm that is self updating and can score any player even if it has never seen them before simply based on their stats.
The model never sees player names, team IDs, or outcome labels from the season it's predicting. Calibrated on the 2024 injury set, evaluated against unseen 2025 outcomes — 72% ROC-AUC, fully reproducible.
We scored every 2025 roster pre-season using only pre-2025 data, then checked how those scores sorted against the injuries that actually happened — out-of-sample, on a season the model never saw.
Full honesty: 72% ranking accuracy still leaves 28% of injured-vs-healthy pairs misranked, and no model can predict a purely random contact injury. We publish the real out-of-sample numbers — recomputed live on the Results tab — not cherry-picked ones.
We fed every notable 2025 injury through INJSUR's pre-season model and asked a simple question: could you have seen it coming? In most cases, the answer was yes.
Across every (2025-injured · healthy) player pair, the model ranked the injured player higher 72% of the time — using only pre-2025 features. Coin flip is 50%; this is a real, reproducible edge.
Players flagged elevated-risk missed time at ~1.7× the league base rate in 2025 — about 1 in 3 flagged players vs ~1 in 5 league-wide. That gap never makes it into the line.
The top 20% of the pre-season risk ranking accounted for 32% of all 2025 missed-game injuries — roughly 1.6× what a random 20% slice would catch. That's the tier to hedge.
78% of 2025 injured players ranked in the top half of the risk distribution before the season — the durable half of the league produced just ~1 in 5 injuries. See the full backtest →
Same features on both plans. $39 for the season, $99 once for lifetime access.
Full access for the entire 2026 NFL season — every team, every player, every week.
Less than one bet for the season.
Pay once — use every NFL season, including every algorithm update and new feature we ship.
Less than the change in your couch for a lifetime.
72% ROC-AUC on the unseen 2025 holdout. Full validation and competitive comparisons available.
See Accuracy & Validation →Lock in your subscription now and get every Pro feature as it ships — no upsells, no add-on fees. Here's what's on the roadmap.
Have ideas for more awesome features? Email us. We'll process them and build them straight into Pro as we build this together.
The same frailty score powers every format — DFS, props, season-long, dynasty. See how the model works and how to apply it to the games you already play.
The players who produce the most points, but carry the highest risk of breaking down. That's the list. That's the whole product. It updates every week.
"I didn't lose because I picked wrong. I lost because nobody was even looking at who was most likely to get hurt. Thirty-five years later, I built the thing I wish I'd had that morning."
Built by an avid fantasy football player and bettor with 35 years in the game. Not a hedge fund, not a chatbot. Read the full story →
Coverage, validation, updates, pricing — everything you need to know, answered in one place.
Type any name, see the frailty score. No account, no card. When you're ready, lock in the full season for $39 — or pay $99 once for lifetime access.
Try 5 FreeINJSUR is a decision-support tool. Past backtest accuracy does not guarantee future results. Please play and wager within your limits.