DEVELOPER DOCS

NOPUNT API

Everything we publish is also exposed as JSON. Permissive CORS so you can call it from any frontend. Use it for dashboards, Discord bots, Twitter automation, fantasy-league widgets, whatever you build, ship it.

ENDPOINTS

REQ / MIN

JSON + CSV

FORMATS

⚠JSON endpoints need no authentication. /api/picks.json returns the full analytical read for every game, pick side, tier, confidence, predicted margin, spread, and moneylines. Pro sessions additionally receive the Edge fields. CSV data exports are Pro only and require an active session.

# Quick start
curl https://www.nopunt.com/api/picks.json
curl "https://www.nopunt.com/api/players.json?pos=RB&limit=10"
curl https://www.nopunt.com/api/power-rankings.json
curl "https://www.nopunt.com/api/fantasy.json?pos=RB"  # Pro only
curl https://www.nopunt.com/api/player-props.json  # Pro only

PICKS

Current and historical game picks

GET/api/picks.jsonCACHE 5 MINUTESTry it →

The most recently published slate, confidence tiers, spreads, model moneylines, and predicted margins. Read slate_state before treating these as upcoming: it is "upcoming" before the first kickoff, "in_progress" during the week, and "final" once the last game has been played. Between the end of one week and Wednesday's new picks, the newest slate is a completed one.

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
game_id	string	Stable nflverse id, e.g. 2025_18_GB_MIN. The join key for every other endpoint and the /matchup URL slug.
home_team	string	Home team abbreviation
away_team	string	Away team abbreviation
pick_side	HOME \| AWAY	Which side pick_team is. REQUIRED to interpret predicted_margin and spread, both of which are home-relative, multiply by -1 when this is AWAY to get the pick's point of view.
kickoff_at	string \| null	ISO 8601 UTC kickoff. Null when the schedule has not published a time yet.
pick_team	string	Team abbreviation the model picked to cover
tier	S \| A+ \| A \| B \| C	Confidence tier. S is highest conviction
confidence	number	Picked-side win probability (0-1)
predicted_margin	number	Projected margin, HOME-RELATIVE: positive = the home team wins by that much, negative = the away team does. NOT relative to pick_team, for an away pick a positive value favours the opponent.
spread	number	Closing line on the SAME home-relative convention: positive = home favoured. Directly comparable to predicted_margin; both are from the home team's point of view.
model_ml	number	Fair, vig-free moneyline implied by the MARKET's spread for the picked side, not by our model. Despite the name it carries no model input: it is spread_line run through a logistic and converted to American odds. Our own number is confidence; convert that if you want our price.
live_ml	number	Best live book moneyline for the pick

EXAMPLE RESPONSE

{
  "season": 2024,
  "week": 12,
  "generated_at": "2024-11-19T14:00:00.000Z",
  "slate_state": "upcoming",
  "picks": [
    {
      "game_id": "2024_12_SF_GB",
      "away_team": "SF",
      "home_team": "GB",
      "pick_team": "SF",
      "pick_side": "AWAY",
      "tier": "S",
      "confidence": 0.83,
      "predicted_margin": -11.5,
      "spread": -8,
      "model_ml": -320,
      "live_ml": -310,
      "kickoff_at": "2024-11-24T18:25:00Z"
    }
  ]
}

GET/api/picks/{season}/{week}.jsonCACHE 24 HOURS

Any historical week with picks and final results joined. Returns 404 if the week has no picks on file.

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
season	2024	Four-digit NFL season year
week	5	Week number 1-18. Regular season only; the model has never produced a postseason pick, so weeks 19+ will 404.

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
result	object \| null	Null if game hasn't finished yet
result.correct	boolean \| null	Whether the picked team won straight up. NULL on a tie, not false. Testing if (!result.correct) books a tie as a loss; test for null first.
result.profit	number \| null	NULL on a tie or an ungraded game. Profit/loss at a 1u flat moneyline stake (1u = $1), priced at the MARKET's spread-implied, vig-free moneyline (see model_ml), not a price any sportsbook offers. Treat it as a comparable series, not achievable P/L.

EXAMPLE RESPONSE

{
  "season": 2024,
  "week": 5,
  "generated_at": "2024-10-09T12:00:00.000Z",
  "slate_state": "final",
  "picks": [
    {
      "game_id": "2024_05_ARI_SF",
      "away_team": "ARI",
      "home_team": "SF",
      "pick_team": "SF",
      "pick_side": "HOME",
      "tier": "A+",
      "confidence": 0.733997,
      "predicted_margin": 8.4,
      "spread": 7,
      "model_ml": -274,
      "live_ml": null,
      "kickoff_at": "2024-10-06T20:05:00+00:00",
      "result": {
        "away_score": 24,
        "home_score": 23,
        "correct": false,
        "profit": -1
      }
    }
  ]
}

ANALYTICS

Performance stats, team records, rankings, and player data

GET/api/stats.jsonCACHE 10 MINUTESTry it →

All-time record + per-tier hit rates + per-season breakdown. The single endpoint for embedding NoPunt performance in dashboards.

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
all_time.accuracy	number	Win rate as decimal (0-1)
all_time.decided_games	number	Games with a winner. total_games minus ties, divide by THIS for a hit rate, not by total_games
all_time.total_profit	number	All-pick flat-1u profit (1u = $1) at spread-implied vig-free prices
by_tier	object	Keyed by tier (S, A+, A, B, C) with wins/total/accuracy/profit. Note the two move independently: accuracy rises with tier, profit does not, S and A+ are near flat while B and C carry the P/L.
by_season	array	Per-season breakdown sorted most recent first

EXAMPLE RESPONSE

{
  "generated_at": "2024-11-19T14:00:00.000Z",
  "all_time": {
    "total_games": 1615,
    "decided_games": 1610,
    "correct": 1070,
    "accuracy": 0.6645962732919255,
    "total_profit": 102.32
  },
  "by_tier": {
    "S": { "total": 86, "wins": 72, "accuracy": 0.8372, "profit": 0.0072 },
    "A+": { "total": 201, "wins": 151, "accuracy": 0.7512, "profit": -2.1298 }
  },
  "by_season": [
    { "season": 2025, "total": 271, "wins": 185, "accuracy": 0.6827, "profit": 25.8468 }
  ],
  "weeks_with_picks": 107
}

GET/api/teams.jsonCACHE 10 MINUTESTry it →

32-team list with NoPunt's pick history per franchise: when we picked them, when we picked against them, combined accuracy.

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
abbr	string	Official NFL team abbreviation
picks_for	number	Total times we picked this team to cover
picks_for_pct	number \| null	Win rate when picking this team
combined_accuracy	number \| null	Overall accuracy in games involving this team

EXAMPLE RESPONSE

{
  "generated_at": "2024-11-19T14:00:00.000Z",
  "teams": [
    {
      "abbr": "KC",
      "slug": "kc",
      "picks_for": 42,
      "picks_for_wins": 28,
      "picks_for_pct": 0.667,
      "picks_against": 18,
      "picks_against_wins": 10,
      "picks_against_pct": 0.556,
      "combined_accuracy": 0.633,
      "last_kickoff": "2024-11-17T20:25:00Z"
    }
  ]
}

GET/api/team/{abbr}CACHE 10 MINUTES

Comprehensive per-team analytics hub: EPA performance, coaching tendencies, playoff simulation odds, and NoPunt pick history.

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
abbr	kc	Team abbreviation (case-insensitive)

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
team.power_rank	number \| null	1-32 ranking by net EPA/play
epa.net_epa_play	number \| null	Net expected points added per play
coaching.pass_rate	number \| null	Overall pass play percentage
coaching.fourth_down_go_rate	number \| null	How often team goes for it on 4th down
playoffs.super_bowl_pct	number \| null	Monte Carlo Super Bowl win probability, 0-1 (0.187 = 18.7%)
picks.for_pct	number \| null	Win rate when NoPunt backs this team

EXAMPLE RESPONSE

{
  "season": 2024,
  "team": { "abbr": "KC", "slug": "kc", "power_rank": 3 },
  "picks": { "for_record": "28-14", "for_pct": 0.667, "total_games": 60 },
  "epa": { "net_epa_play": 0.098, "off_epa_play": 0.065, "def_epa_play": -0.033 },
  "coaching": { "pass_rate": 0.617, "fourth_down_go_rate": 0.167, "red_zone_td_rate": 0.792 },
  "playoffs": { "playoff_pct": 0.942, "super_bowl_pct": 0.187 }
}

GET/api/power-rankings.jsonCACHE 1 HOURTry it →

32 teams ranked by composite EPA. Each row includes off_epa_play, def_epa_play, net_epa_play, success rates, and PPG.

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
season	2024	Optional, defaults to most recent backfilled season

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
rank	number	1-32 ranking by net EPA/play
net_epa_play	number	Offense EPA minus defense EPA per play
off_epa_play	number	Expected points added per offensive play
def_epa_play	number	Expected points added per defensive play (lower is better)
off_success	number \| null	Offensive success rate (positive EPA plays / total)
points_for	number \| null	Total points scored

EXAMPLE RESPONSE

{
  "season": 2024,
  "generated_at": "2024-11-19T14:00:00.000Z",
  "rankings": [
    {
      "rank": 1,
      "team": "BAL",
      "slug": "bal",
      "net_epa_play": 0.142,
      "off_epa_play": 0.098,
      "def_epa_play": -0.044,
      "off_success": 0.48,
      "def_success": 0.38,
      "points_for": 312,
      "points_against": 218,
      "games_played": 11
    }
  ]
}

GET/api/players.jsonCACHE 10 MINUTESTry it →

Top N stat leaders by position for the current season (or a specified week). Supports passing/rushing/receiving/PPR.

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
view	season	season (default) or week
pos	QB	QB, RB, WR, or TE
stat	passing_yards	passing_yards \| rushing_yards \| receiving_yards \| fantasy_points_ppr \| etc
limit	20	1-50, default 20
season	2024	Optional, defaults to most recent season with data

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
leaders	array	Ranked list of player stat rows
leaders[].player_name	string	Player display name
leaders[].team	string	Current team abbreviation
leaders[].stat_total	number	Season view only: total for the requested stat (the value rows are ranked by). The week view omits stat_total and instead returns the raw per-game stat columns (passing_yards, rushing_yards, ...) for that week.
leaders[].games_played	number	Games the player has data for this season (season view only)

EXAMPLE RESPONSE

{
  "season": 2024,
  "view": "season",
  "pos": "QB",
  "stat": "passing_yards",
  "limit": 3,
  "leaders": [
    {
      "player_name": "Joe Burrow",
      "team": "CIN",
      "position": "QB",
      "stat_total": 4918,
      "games_played": 17,
      "passing_yards": null,
      "passing_tds": null,
      "fantasy_points_ppr": null
    }
  ]
}

GET/api/nextgen.jsonCACHE 10 MINUTESTry it →

Next Gen Stats leaderboard (public, CORS-enabled). Passing/rushing/receiving process metrics. week=0 (default) is the season aggregate; pass a week for a weekly board.

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
stat_type	passing	passing (default), rushing, or receiving
metric	avg_time_to_throw	Optional, defaults to the primary metric for the stat_type
week	0	Optional, 0 (default) = season aggregate, or a week number
season	2025	Optional, defaults to most recent season with data
limit	20	Optional, 1–50, default 20

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
leaders	array	Leaderboard rows ranked by the metric; each row also carries the requested metric as its own key (e.g. avg_time_to_throw)
leaders[].player_name	string	Player display name
leaders[].team	string	Team abbreviation
leaders[].position	string	Player position

EXAMPLE RESPONSE

{
  "season": 2025,
  "week": 0,
  "stat_type": "passing",
  "metric": "avg_time_to_throw",
  "count": 20,
  "leaders": [
    { "player_name": "J.Goff", "team": "det", "position": "QB", "avg_time_to_throw": 2.61 }
  ]
}

GET/api/advanced-stats.jsonCACHE 10 MINUTESTry it →

Pro-Football-Reference advanced-stat leaderboard (public, CORS-enabled). Season-level leaders for a chosen advanced metric (pressure, drops, YBC/YAC, coverage, ...).

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
metric	yards_per_target	Metric id (default yards_per_target)
season	2025	Optional, defaults to most recent season with data
limit	20	Optional, 1–50, default 20

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
label	string	Human-readable metric label
leaders	array	Leaderboard rows, ranked by the metric
leaders[].player_name	string	Player display name
leaders[].team	string	Team abbreviation
leaders[].games	number	Games played
leaders[].value	number	The metric value (2dp)

EXAMPLE RESPONSE

{
  "season": 2025,
  "stat_type": "receiving",
  "metric": "yards_per_target",
  "label": "Yards / Target",
  "count": 20,
  "leaders": [
    { "player_name": "N.Collins", "team": "hou", "games": 12, "value": 11.4 }
  ]
}

GET/api/game-wp.jsonCACHE 1 HOURTry it →

In-game win-probability data (public, CORS-enabled). Two modes: pass game_id for a single game's WP curve; omit it (optionally with season) for that season's most dramatic games. SOURCE: the curve is nflfastR's play-by-play home_wp, not a NoPunt model output. Our own numbers are the pregame ones on /api/picks.json.

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
game_id	2025_01_BUF_MIA	Single-game mode, returns that game's WP curve
season	2025	List mode, most dramatic games of the season (ignored when game_id is set)

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
games	array	List mode: the season's most dramatic games. game_id mode instead returns a single game with wp_curve = [[game_seconds_remaining, home_wp], ...]
games[].matchup	string	"AWAY @ HOME"
games[].drama_score	number	Composite drama rating
games[].wp_swings	number	Count of lead-probability swings (the DB column is wp_flips)

EXAMPLE RESPONSE

{
  "season": 2025,
  "count": 12,
  "games": [
    { "game_id": "2025_04_BUF_NE", "matchup": "BUF @ NE", "week": 4, "drama_score": 88, "wp_swings": 6 }
  ]
}

GET/api/fantasy.jsonCACHE NOT CACHED (PRO-GATED, PER-REQUEST)Try it →

PRO ONLY (403 without an active subscription). PPR fantasy projections for a week: projected points, position rank, the player's 4-week rolling average, the opponent's defensive rank against the position, and a confidence grade. Defaults to the latest week with projections.

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
pos	RB	Optional, QB, RB, WR, or TE. Omit for all positions.
week	18	Optional, defaults to the latest week with projections
season	2025	Optional, defaults to most recent season with data

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
projections	array	Projection rows, ranked by position_rank
projections[].player_name	string	Player display name
projections[].projected_points	number	Projected PPR points
projections[].position_rank	number	Rank within position for the week
projections[].rolling_avg_4	number \| null	The player's PPR average over their last 4 games
projections[].opponent_def_rank	number \| null	Opponent's rank against this position, 1 = toughest
projections[].confidence	string	Projection confidence grade

EXAMPLE RESPONSE

{
  "season": 2025,
  "week": 18,
  "pos": "RB",
  "count": 140,
  "projections": [
    {
      "id": 48211,
      "season": 2025,
      "week": 18,
      "player_id": "00-0038542",
      "player_name": "B.Robinson",
      "position": "RB",
      "team": "atl",
      "opponent": "no",
      "projected_points": 30.78,
      "position_rank": 1,
      "rolling_avg_4": 21.4,
      "opponent_def_rank": 27,
      "confidence": "A"
    }
  ]
}

GET/api/player-props.jsonCACHE NOT CACHED (PRO-GATED, PER-REQUEST)Try it →

PRO ONLY (403 without an active subscription). Player prop projections with floor/ceiling and matchup context. Default view returns the top high-confidence model edges; view=all returns every projection grouped by player.

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
view	edges	edges (default, top A-confidence favorable matchups) or all
pos	WR	Optional, QB, RB, WR, or TE (view=all only)
week	18	Optional, defaults to the latest week with projections
season	2025	Optional, defaults to most recent season with data

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
edges	array	view=edges: top high-confidence prop projections by matchup strength
players	array	view=all: projections grouped by player, each with a stats[] array
edges[].stat_type	string	Stat projected (receptions, rushing_yards, ...)
edges[].projected	number	Point projection for the stat
edges[].floor / .ceiling	number	Confidence-interval bounds

EXAMPLE RESPONSE

{
  "season": 2025,
  "week": 18,
  "view": "edges",
  "count": 20,
  "edges": [
    {
      "player_name": "N.Collins",
      "position": "WR",
      "team": "hou",
      "opponent": "ind",
      "stat_type": "receptions",
      "projected": 3.7,
      "floor": 2.5,
      "ceiling": 4.9,
      "confidence": "A"
    }
  ]
}

DATA EXPORTS

Downloadable CSV files for offline analysis

GET/api/exports/predictions.csvCACHE 24 HOURSPRO

Full pick history with joined results as a downloadable CSV. Includes every pick, tier, confidence, spread, moneyline, final score, and W/L outcome. Pro only, requires an active Pro session.

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
game_id	string	Unique game identifier (e.g. 2024_12_SF_GB)
tier	S \| A+ \| A \| B \| C	Confidence tier
correct	boolean \| null	Whether the pick won straight up. NULL on a tie, not false, and there are 5 in the ledger (e.g. 2022_01_IND_HOU 20-20). Testing if (!correct) books a tie as a loss; test for null first. Ties are excluded from accuracy denominators, which is what decided_games is for.
profit	number \| null	NULL on a tie or an ungraded game. Profit/loss at a 1u flat moneyline stake, priced at the MARKET's spread-implied, vig-free moneyline (see model_ml), not a book price. See /docs field notes.

EXAMPLE RESPONSE

game_id,season,week,home_team,away_team,pick_team,...,home_score,away_score,correct,profit,spread_implied_ml
2024_12_SF_GB,2024,12,GB,SF,GB,...,38,10,true,0.3922,-290

GET/api/exports/player-stats.csvCACHE 24 HOURSPRO

Per-player weekly stat lines, passing, rushing, receiving, and fantasy points. 2020-present. Pro only, requires an active Pro session.

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
player_id	string	NFL GSIS player ID
player_name	string	Abbreviated as the source feed provides it, S.Barkley, not Saquon Barkley
team	string	LOWERCASE here (phi), unlike the UPPERCASE codes in the picks and team-stats endpoints. Normalize case before joining across endpoints.
fantasy_points_ppr	number	PPR fantasy points for the week

EXAMPLE RESPONSE

player_id,player_name,position,team,season,week,...,fantasy_points_ppr
00-0034844,S.Barkley,RB,phi,2024,12,...,46.2

GET/api/exports/team-stats.csvCACHE 24 HOURSPRO

Per-team season EPA composites, offensive, defensive, net EPA per play, pass/rush splits, and point totals. Pro only, requires an active Pro session.

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
net_epa_play	number	Net EPA per play (offense - defense)
off_epa_play	number	Offensive EPA per play
def_epa_play	number	EPA per play ALLOWED by this defense, signed from the offense point of view, lower is better, and a good defense is often still positive. net_epa_play = off_epa_play - def_epa_play.

EXAMPLE RESPONSE

team,season,games_played,off_epa_play,def_epa_play,net_epa_play,...
KC,2024,17,0.07,0.0112,0.0588,...

FIELD	TYPE	DESCRIPTION
status	"ok" \| "ok_offseason_empty" \| "stale" \| "degraded"	Overall service health (only field returned to anonymous callers). ok/ok_offseason_empty → 200; stale/degraded → 503.

FIELD	TYPE	DESCRIPTION
channel/item	XML	One item per completed game with title, description, pubDate, guid

NOPUNT API

ENDPOINTS

REQ / MIN

JSON + CSV

FORMATS

⚠JSON endpoints need no authentication. /api/picks.json returns the full analytical read for every game, pick side, tier, confidence, predicted margin, spread, and moneylines. Pro sessions additionally receive the Edge fields. CSV data exports are Pro only and require an active session.

# Quick start
curl https://www.nopunt.com/api/picks.json
curl "https://www.nopunt.com/api/players.json?pos=RB&limit=10"
curl https://www.nopunt.com/api/power-rankings.json
curl "https://www.nopunt.com/api/fantasy.json?pos=RB"  # Pro only
curl https://www.nopunt.com/api/player-props.json  # Pro only

PICKS

Current and historical game picks

GET/api/picks.jsonCACHE 5 MINUTESTry it →

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
game_id	string	Stable nflverse id, e.g. 2025_18_GB_MIN. The join key for every other endpoint and the /matchup URL slug.
home_team	string	Home team abbreviation
away_team	string	Away team abbreviation
pick_side	HOME \| AWAY	Which side pick_team is. REQUIRED to interpret predicted_margin and spread, both of which are home-relative, multiply by -1 when this is AWAY to get the pick's point of view.
kickoff_at	string \| null	ISO 8601 UTC kickoff. Null when the schedule has not published a time yet.
pick_team	string	Team abbreviation the model picked to cover
tier	S \| A+ \| A \| B \| C	Confidence tier. S is highest conviction
confidence	number	Picked-side win probability (0-1)
predicted_margin	number	Projected margin, HOME-RELATIVE: positive = the home team wins by that much, negative = the away team does. NOT relative to pick_team, for an away pick a positive value favours the opponent.
spread	number	Closing line on the SAME home-relative convention: positive = home favoured. Directly comparable to predicted_margin; both are from the home team's point of view.
model_ml	number	Fair, vig-free moneyline implied by the MARKET's spread for the picked side, not by our model. Despite the name it carries no model input: it is spread_line run through a logistic and converted to American odds. Our own number is confidence; convert that if you want our price.
live_ml	number	Best live book moneyline for the pick

EXAMPLE RESPONSE

{
  "season": 2024,
  "week": 12,
  "generated_at": "2024-11-19T14:00:00.000Z",
  "slate_state": "upcoming",
  "picks": [
    {
      "game_id": "2024_12_SF_GB",
      "away_team": "SF",
      "home_team": "GB",
      "pick_team": "SF",
      "pick_side": "AWAY",
      "tier": "S",
      "confidence": 0.83,
      "predicted_margin": -11.5,
      "spread": -8,
      "model_ml": -320,
      "live_ml": -310,
      "kickoff_at": "2024-11-24T18:25:00Z"
    }
  ]
}

GET/api/picks/{season}/{week}.jsonCACHE 24 HOURS

Any historical week with picks and final results joined. Returns 404 if the week has no picks on file.

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
season	2024	Four-digit NFL season year
week	5	Week number 1-18. Regular season only; the model has never produced a postseason pick, so weeks 19+ will 404.

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
result	object \| null	Null if game hasn't finished yet
result.correct	boolean \| null	Whether the picked team won straight up. NULL on a tie, not false. Testing if (!result.correct) books a tie as a loss; test for null first.
result.profit	number \| null	NULL on a tie or an ungraded game. Profit/loss at a 1u flat moneyline stake (1u = $1), priced at the MARKET's spread-implied, vig-free moneyline (see model_ml), not a price any sportsbook offers. Treat it as a comparable series, not achievable P/L.

EXAMPLE RESPONSE

{
  "season": 2024,
  "week": 5,
  "generated_at": "2024-10-09T12:00:00.000Z",
  "slate_state": "final",
  "picks": [
    {
      "game_id": "2024_05_ARI_SF",
      "away_team": "ARI",
      "home_team": "SF",
      "pick_team": "SF",
      "pick_side": "HOME",
      "tier": "A+",
      "confidence": 0.733997,
      "predicted_margin": 8.4,
      "spread": 7,
      "model_ml": -274,
      "live_ml": null,
      "kickoff_at": "2024-10-06T20:05:00+00:00",
      "result": {
        "away_score": 24,
        "home_score": 23,
        "correct": false,
        "profit": -1
      }
    }
  ]
}

ANALYTICS

Performance stats, team records, rankings, and player data

GET/api/stats.jsonCACHE 10 MINUTESTry it →

All-time record + per-tier hit rates + per-season breakdown. The single endpoint for embedding NoPunt performance in dashboards.

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
all_time.accuracy	number	Win rate as decimal (0-1)
all_time.decided_games	number	Games with a winner. total_games minus ties, divide by THIS for a hit rate, not by total_games
all_time.total_profit	number	All-pick flat-1u profit (1u = $1) at spread-implied vig-free prices
by_tier	object	Keyed by tier (S, A+, A, B, C) with wins/total/accuracy/profit. Note the two move independently: accuracy rises with tier, profit does not, S and A+ are near flat while B and C carry the P/L.
by_season	array	Per-season breakdown sorted most recent first

EXAMPLE RESPONSE

{
  "generated_at": "2024-11-19T14:00:00.000Z",
  "all_time": {
    "total_games": 1615,
    "decided_games": 1610,
    "correct": 1070,
    "accuracy": 0.6645962732919255,
    "total_profit": 102.32
  },
  "by_tier": {
    "S": { "total": 86, "wins": 72, "accuracy": 0.8372, "profit": 0.0072 },
    "A+": { "total": 201, "wins": 151, "accuracy": 0.7512, "profit": -2.1298 }
  },
  "by_season": [
    { "season": 2025, "total": 271, "wins": 185, "accuracy": 0.6827, "profit": 25.8468 }
  ],
  "weeks_with_picks": 107
}

GET/api/teams.jsonCACHE 10 MINUTESTry it →

32-team list with NoPunt's pick history per franchise: when we picked them, when we picked against them, combined accuracy.

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
abbr	string	Official NFL team abbreviation
picks_for	number	Total times we picked this team to cover
picks_for_pct	number \| null	Win rate when picking this team
combined_accuracy	number \| null	Overall accuracy in games involving this team

EXAMPLE RESPONSE

{
  "generated_at": "2024-11-19T14:00:00.000Z",
  "teams": [
    {
      "abbr": "KC",
      "slug": "kc",
      "picks_for": 42,
      "picks_for_wins": 28,
      "picks_for_pct": 0.667,
      "picks_against": 18,
      "picks_against_wins": 10,
      "picks_against_pct": 0.556,
      "combined_accuracy": 0.633,
      "last_kickoff": "2024-11-17T20:25:00Z"
    }
  ]
}

GET/api/team/{abbr}CACHE 10 MINUTES

Comprehensive per-team analytics hub: EPA performance, coaching tendencies, playoff simulation odds, and NoPunt pick history.

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
abbr	kc	Team abbreviation (case-insensitive)

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
team.power_rank	number \| null	1-32 ranking by net EPA/play
epa.net_epa_play	number \| null	Net expected points added per play
coaching.pass_rate	number \| null	Overall pass play percentage
coaching.fourth_down_go_rate	number \| null	How often team goes for it on 4th down
playoffs.super_bowl_pct	number \| null	Monte Carlo Super Bowl win probability, 0-1 (0.187 = 18.7%)
picks.for_pct	number \| null	Win rate when NoPunt backs this team

EXAMPLE RESPONSE

{
  "season": 2024,
  "team": { "abbr": "KC", "slug": "kc", "power_rank": 3 },
  "picks": { "for_record": "28-14", "for_pct": 0.667, "total_games": 60 },
  "epa": { "net_epa_play": 0.098, "off_epa_play": 0.065, "def_epa_play": -0.033 },
  "coaching": { "pass_rate": 0.617, "fourth_down_go_rate": 0.167, "red_zone_td_rate": 0.792 },
  "playoffs": { "playoff_pct": 0.942, "super_bowl_pct": 0.187 }
}

GET/api/power-rankings.jsonCACHE 1 HOURTry it →

32 teams ranked by composite EPA. Each row includes off_epa_play, def_epa_play, net_epa_play, success rates, and PPG.

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
season	2024	Optional, defaults to most recent backfilled season

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
rank	number	1-32 ranking by net EPA/play
net_epa_play	number	Offense EPA minus defense EPA per play
off_epa_play	number	Expected points added per offensive play
def_epa_play	number	Expected points added per defensive play (lower is better)
off_success	number \| null	Offensive success rate (positive EPA plays / total)
points_for	number \| null	Total points scored

EXAMPLE RESPONSE

{
  "season": 2024,
  "generated_at": "2024-11-19T14:00:00.000Z",
  "rankings": [
    {
      "rank": 1,
      "team": "BAL",
      "slug": "bal",
      "net_epa_play": 0.142,
      "off_epa_play": 0.098,
      "def_epa_play": -0.044,
      "off_success": 0.48,
      "def_success": 0.38,
      "points_for": 312,
      "points_against": 218,
      "games_played": 11
    }
  ]
}

GET/api/players.jsonCACHE 10 MINUTESTry it →

Top N stat leaders by position for the current season (or a specified week). Supports passing/rushing/receiving/PPR.

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
view	season	season (default) or week
pos	QB	QB, RB, WR, or TE
stat	passing_yards	passing_yards \| rushing_yards \| receiving_yards \| fantasy_points_ppr \| etc
limit	20	1-50, default 20
season	2024	Optional, defaults to most recent season with data

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
leaders	array	Ranked list of player stat rows
leaders[].player_name	string	Player display name
leaders[].team	string	Current team abbreviation
leaders[].stat_total	number	Season view only: total for the requested stat (the value rows are ranked by). The week view omits stat_total and instead returns the raw per-game stat columns (passing_yards, rushing_yards, ...) for that week.
leaders[].games_played	number	Games the player has data for this season (season view only)

EXAMPLE RESPONSE

{
  "season": 2024,
  "view": "season",
  "pos": "QB",
  "stat": "passing_yards",
  "limit": 3,
  "leaders": [
    {
      "player_name": "Joe Burrow",
      "team": "CIN",
      "position": "QB",
      "stat_total": 4918,
      "games_played": 17,
      "passing_yards": null,
      "passing_tds": null,
      "fantasy_points_ppr": null
    }
  ]
}

GET/api/nextgen.jsonCACHE 10 MINUTESTry it →

Next Gen Stats leaderboard (public, CORS-enabled). Passing/rushing/receiving process metrics. week=0 (default) is the season aggregate; pass a week for a weekly board.

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
stat_type	passing	passing (default), rushing, or receiving
metric	avg_time_to_throw	Optional, defaults to the primary metric for the stat_type
week	0	Optional, 0 (default) = season aggregate, or a week number
season	2025	Optional, defaults to most recent season with data
limit	20	Optional, 1–50, default 20

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
leaders	array	Leaderboard rows ranked by the metric; each row also carries the requested metric as its own key (e.g. avg_time_to_throw)
leaders[].player_name	string	Player display name
leaders[].team	string	Team abbreviation
leaders[].position	string	Player position

EXAMPLE RESPONSE

{
  "season": 2025,
  "week": 0,
  "stat_type": "passing",
  "metric": "avg_time_to_throw",
  "count": 20,
  "leaders": [
    { "player_name": "J.Goff", "team": "det", "position": "QB", "avg_time_to_throw": 2.61 }
  ]
}

GET/api/advanced-stats.jsonCACHE 10 MINUTESTry it →

Pro-Football-Reference advanced-stat leaderboard (public, CORS-enabled). Season-level leaders for a chosen advanced metric (pressure, drops, YBC/YAC, coverage, ...).

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
metric	yards_per_target	Metric id (default yards_per_target)
season	2025	Optional, defaults to most recent season with data
limit	20	Optional, 1–50, default 20

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
label	string	Human-readable metric label
leaders	array	Leaderboard rows, ranked by the metric
leaders[].player_name	string	Player display name
leaders[].team	string	Team abbreviation
leaders[].games	number	Games played
leaders[].value	number	The metric value (2dp)

EXAMPLE RESPONSE

{
  "season": 2025,
  "stat_type": "receiving",
  "metric": "yards_per_target",
  "label": "Yards / Target",
  "count": 20,
  "leaders": [
    { "player_name": "N.Collins", "team": "hou", "games": 12, "value": 11.4 }
  ]
}

GET/api/game-wp.jsonCACHE 1 HOURTry it →

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
game_id	2025_01_BUF_MIA	Single-game mode, returns that game's WP curve
season	2025	List mode, most dramatic games of the season (ignored when game_id is set)

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
games	array	List mode: the season's most dramatic games. game_id mode instead returns a single game with wp_curve = [[game_seconds_remaining, home_wp], ...]
games[].matchup	string	"AWAY @ HOME"
games[].drama_score	number	Composite drama rating
games[].wp_swings	number	Count of lead-probability swings (the DB column is wp_flips)

EXAMPLE RESPONSE

{
  "season": 2025,
  "count": 12,
  "games": [
    { "game_id": "2025_04_BUF_NE", "matchup": "BUF @ NE", "week": 4, "drama_score": 88, "wp_swings": 6 }
  ]
}

GET/api/fantasy.jsonCACHE NOT CACHED (PRO-GATED, PER-REQUEST)Try it →

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
pos	RB	Optional, QB, RB, WR, or TE. Omit for all positions.
week	18	Optional, defaults to the latest week with projections
season	2025	Optional, defaults to most recent season with data

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
projections	array	Projection rows, ranked by position_rank
projections[].player_name	string	Player display name
projections[].projected_points	number	Projected PPR points
projections[].position_rank	number	Rank within position for the week
projections[].rolling_avg_4	number \| null	The player's PPR average over their last 4 games
projections[].opponent_def_rank	number \| null	Opponent's rank against this position, 1 = toughest
projections[].confidence	string	Projection confidence grade

EXAMPLE RESPONSE

{
  "season": 2025,
  "week": 18,
  "pos": "RB",
  "count": 140,
  "projections": [
    {
      "id": 48211,
      "season": 2025,
      "week": 18,
      "player_id": "00-0038542",
      "player_name": "B.Robinson",
      "position": "RB",
      "team": "atl",
      "opponent": "no",
      "projected_points": 30.78,
      "position_rank": 1,
      "rolling_avg_4": 21.4,
      "opponent_def_rank": 27,
      "confidence": "A"
    }
  ]
}

GET/api/player-props.jsonCACHE NOT CACHED (PRO-GATED, PER-REQUEST)Try it →

PARAMETERS

PARAM	EXAMPLE	DESCRIPTION
view	edges	edges (default, top A-confidence favorable matchups) or all
pos	WR	Optional, QB, RB, WR, or TE (view=all only)
week	18	Optional, defaults to the latest week with projections
season	2025	Optional, defaults to most recent season with data

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
edges	array	view=edges: top high-confidence prop projections by matchup strength
players	array	view=all: projections grouped by player, each with a stats[] array
edges[].stat_type	string	Stat projected (receptions, rushing_yards, ...)
edges[].projected	number	Point projection for the stat
edges[].floor / .ceiling	number	Confidence-interval bounds

EXAMPLE RESPONSE

{
  "season": 2025,
  "week": 18,
  "view": "edges",
  "count": 20,
  "edges": [
    {
      "player_name": "N.Collins",
      "position": "WR",
      "team": "hou",
      "opponent": "ind",
      "stat_type": "receptions",
      "projected": 3.7,
      "floor": 2.5,
      "ceiling": 4.9,
      "confidence": "A"
    }
  ]
}

DATA EXPORTS

Downloadable CSV files for offline analysis

GET/api/exports/predictions.csvCACHE 24 HOURSPRO

Full pick history with joined results as a downloadable CSV. Includes every pick, tier, confidence, spread, moneyline, final score, and W/L outcome. Pro only, requires an active Pro session.

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
game_id	string	Unique game identifier (e.g. 2024_12_SF_GB)
tier	S \| A+ \| A \| B \| C	Confidence tier
correct	boolean \| null	Whether the pick won straight up. NULL on a tie, not false, and there are 5 in the ledger (e.g. 2022_01_IND_HOU 20-20). Testing if (!correct) books a tie as a loss; test for null first. Ties are excluded from accuracy denominators, which is what decided_games is for.
profit	number \| null	NULL on a tie or an ungraded game. Profit/loss at a 1u flat moneyline stake, priced at the MARKET's spread-implied, vig-free moneyline (see model_ml), not a book price. See /docs field notes.

EXAMPLE RESPONSE

game_id,season,week,home_team,away_team,pick_team,...,home_score,away_score,correct,profit,spread_implied_ml
2024_12_SF_GB,2024,12,GB,SF,GB,...,38,10,true,0.3922,-290

GET/api/exports/player-stats.csvCACHE 24 HOURSPRO

Per-player weekly stat lines, passing, rushing, receiving, and fantasy points. 2020-present. Pro only, requires an active Pro session.

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
player_id	string	NFL GSIS player ID
player_name	string	Abbreviated as the source feed provides it, S.Barkley, not Saquon Barkley
team	string	LOWERCASE here (phi), unlike the UPPERCASE codes in the picks and team-stats endpoints. Normalize case before joining across endpoints.
fantasy_points_ppr	number	PPR fantasy points for the week

EXAMPLE RESPONSE

player_id,player_name,position,team,season,week,...,fantasy_points_ppr
00-0034844,S.Barkley,RB,phi,2024,12,...,46.2

GET/api/exports/team-stats.csvCACHE 24 HOURSPRO

Per-team season EPA composites, offensive, defensive, net EPA per play, pass/rush splits, and point totals. Pro only, requires an active Pro session.

RESPONSE FIELDS

FIELD	TYPE	DESCRIPTION
net_epa_play	number	Net EPA per play (offense - defense)
off_epa_play	number	Offensive EPA per play
def_epa_play	number	EPA per play ALLOWED by this defense, signed from the offense point of view, lower is better, and a good defense is often still positive. net_epa_play = off_epa_play - def_epa_play.

EXAMPLE RESPONSE

team,season,games_played,off_epa_play,def_epa_play,net_epa_play,...
KC,2024,17,0.07,0.0112,0.0588,...

NOPUNT API

PICKS

ANALYTICS

DATA EXPORTS

META

ATTRIBUTION

NOPUNT API

PICKS

ANALYTICS

DATA EXPORTS

META

ATTRIBUTION