State of the art NFL projections. Powered by data + AI.
Built for elite DFS sharps and optimizers. Real-time market anchoring, machine-learning ownership dispersion, 9-quantile outcome modeling, and scheme-level DvP. No black box.
GET /api/v1/dfs/slates/96108/projections
{
"player": {
"name": "Jahmyr Gibbs",
"position": "RB",
"salary": 8000
},
"projection": {
"fantasy_pts": "20.29",
"floor": "9.64",
"ceiling": "32.26",
"value": "2.54"
},
"ownership": {
"ownership_pct": "13.68",
"volatility": "4.27",
"chalk_fragility": "0.31"
},
"dvp": {
"archetype": "Pass-Catching Back",
"rating": "0.8866",
"matchup": "Tough"
},
"provenance": {
"update_explanation": "Projection +0.05 FPTS (20.24 -> 20.29): Opponent defensive front injuries (Cameron Jordan (DE)) improve rushing efficiency.",
"update_reason": {
"primary_code": "DEFENSE_DEGRADED"
}
}
}
Data + AI architecture. Nothing held back.
State of the art, in DFS terms.
Explainable AI (no black box).
You always know why a projection moved. When defensive front injuries occur, the engine logs the exact delta and reason code.
The box score always adds up.
Receiving yards sum to passing yards, carries to team attempts. Simplex mass conservation guarantees physical feasibility.
Ownership dispersion & fragility.
Point ownership, exposure volatility, and a fragility index (0-1). Know which chalk survives adverse scripts and which will crater.
Monotonic 9-quantile probability.
Strict outcome distributions (p05 to p95) with p15 floor and p85 ceiling. Target tournament ceiling without sacrificing cash floor.
Every dimension on a player card.
- ✓Points, floor (p15), ceiling (p85), nine percentiles p05–p95
- ✓Full opportunity tree: snaps, snap %, carry %, target %, catch rate
- ✓Ownership dispersion: point estimate, volatility, and chalk fragility
- ✓Scheme-level DvP: archetype, rank, rating, matchup grade, and explanation
- ✓Defensive scheme metrics: shadow CB, DB injuries, funnel tilt, injury boost
- ✓Provenance tracking: last update time, reason code, and delta explanation
- ✓De-vigged player props: anytime TD, first/last TD, rushing/receiving lines
- ✓Team context on every row: Vegas spread and total, projected team snaps
Everything we project, one real player.
Jahmyr Gibbs on DraftKings Slate 96108 (Week 1 2026), exactly as returned by the live API on 2026-09-10. Structured into 8 clear modules with full update provenance.
{
"slate": {
"id": 96108,
"external_id": "151307",
"operator": "DraftKings"
},
"projections": [
{
// 1. Identity & automated scouting
"player": {
"id": 15707370,
"slate_player_id": 15707370,
"player_id": 2478,
"name": "Jahmyr Gibbs",
"position": "RB",
"team": "DET",
"opponent": "NO",
"salary": 8000,
"status": "Active",
"game_id": 43407,
"slate_date": "2026-09-13",
"player_game_note": "Expected to handle a clear feature back workload with elevated snap share."
},
// 2. Points, value, and monotonic 9-quantile distribution
"projection": {
"fantasy_pts": "20.29",
"floor": "9.6367",
"ceiling": "32.2625",
"std_dev": "10.9250",
"value": "2.54",
"quantiles": { "p05": 6.7082, "p10": 8.3087, "p50": 17.8353, "p90": 36.1845, "p95": 40.881 }
},
// 3. Biologically constrained opportunity chain
"opportunity": {
"snap_pct": "77.45",
"projected_snaps": "49.97",
"projected_team_snaps": "64.52",
"carries": "19.2181",
"targets": "6.2985",
"receptions": "4.4102",
"catch_rate": "70.02"
},
// 4. Exact 27-component fantasy scoring breakdown
"breakdown": {
"rushing_yds": "83.6275",
"rushing_tds": "0.6347",
"rushing_yds_fpts": "8.36",
"receiving_yds": "30.5094",
"receiving_tds": "0.1274",
"receiving_yds_fpts": "3.05",
"reception_fpts": "4.41"
},
// 5. Ownership dispersion & chalk fragility
"ownership": {
"projected_ownership": "0.1368",
"ownership_pct": "13.68",
"position_ownership_pct": "5.61",
"volatility": "4.27",
"chalk_fragility": "0.31"
},
// 6. Scheme-level DvP and defensive funnel metrics
"dvp": {
"archetype": "Pass-Catching Back",
"rank": 10,
"rating": "0.8866",
"matchup": "Tough",
"explanation": "Opponent defensive front injuries (Cameron Jordan (DE)) improve rushing efficiency",
"scheme_metrics": {
"opponent": "NO",
"composite": 0.8866,
"shadow_cb": null,
"db_injuries": [],
"funnel_tilt": 1.0,
"scheme_mult": 1.0,
"injury_boost": 1.05,
"pressure_mult": 1.0
}
},
// 7. Sharp de-vigged betting market anchors
"vegas_odds": {
"last_td": 375,
"first_td": 280,
"anytime_td": -322,
"receptions": 4.5,
"rushing_yds": 83.5,
"receiving_yds": 30.5,
"rushing_attempts": 19.5
},
// 8. Explainable AI provenance & audit trail
"provenance": {
"last_update_time": "2026-09-10T16:56:43.455876Z",
"update_explanation": "Projection +0.05 FPTS (20.24 -> 20.29): Opponent defensive front injuries (Cameron Jordan (DE)) improve rushing efficiency.",
"update_reason": {
"primary_code": "DEFENSE_DEGRADED",
"factors": {
"defense_matchup": {
"dvp_rank": 10,
"dvp_rating": 0.8866,
"dvp_archetype": "Pass-Catching Back"
}
}
},
"update_history": [
{
"id": "chg_20260910T165643_15707370",
"previous_value": { "projection": 20.24, "floor": 9.59, "ceiling": 32.21, "projected_ownership": 0.1366, "projected_snap_share": 0.7745 },
"current_value": { "projection": 20.29, "floor": 9.64, "ceiling": 32.26, "projected_ownership": 0.1366, "projected_snap_share": 0.7745 },
"reason": { "primary_code": "DEFENSE_DEGRADED" },
"timestamp": "2026-09-10T16:56:43.455876+00:00",
"explanation": "Projection +0.05 FPTS (20.24 -> 20.29): Opponent defensive front injuries (Cameron Jordan (DE)) improve rushing efficiency."
}
]
}
}
]
}
One call to the slate.
Plain REST, JSON out. The slate endpoint delivers every player with provenance and quantiles.
API=https://api.stat-api.com/api/v1
AUTH="Authorization: Bearer YOUR_PRO_API_KEY"
# one slate: player projections — points, floor, ceiling, ownership, value
curl "$API/dfs/slates/96108/projections" -H "$AUTH"
# → { "slate": { "id": 96108, "operator": "DraftKings", ... },
# "projections": [ { "player_name": "Jahmyr Gibbs", "projected_fpts": 20.29, ... } ] }
Projection questions, answered.
What makes this AI explainable rather than a black box?
Every projection shift logs an ISO timestamp, exact delta, primary reason code (like DEFENSE_DEGRADED), and plain-English explanation of why the number moved.
What is chalk fragility and how do I use it in DFS?
Chalk fragility (0.0 to 1.0) measures how sensitive high-owned players are to adverse game flow. Fragile chalk (>0.40) busts often; stable chalk (<0.20) holds up.
How are the 9 quantiles calculated?
We run 2,000 whole-game Monte Carlo draws preserving actual player covariance, producing a strict monotonic distribution where p15 is floor and p85 is ceiling.
Which operators and contest formats are supported?
DraftKings and FanDuel, each calibrated to its own salaries, positions, roster slots, and scoring rules across classic and showdown slates.
How frequently are projections updated ahead of lock?
Continuously refreshed as practice reports, depth chart tags, weather azimuth vectors, and sharp betting prop markets move right up to lock.
What plan do I need, and can other people use it?
Pro plan — $3.20 per day ($99/mo). Pro is a single-user license: projections, optimizer integration, and private modeling for your own use. You may not publish the projections in any shape or form. Elite ($199/mo) adds more records and a commercial license for your own site or app, where the projections may feed your own numbers but may not show as they are. On Pro, any other person — a teammate, an internal tool, or an app's users, paid or free — and any distribution need our Enterprise License. No plan permits reselling the data or the projections as a feed.
Get the AI edge the sharps kept hidden.
Included in Pro at $3.20 per day — $99/mo.