All your daily fantasy data and tools in one place.
We are dedicated to helping daily fantasy players succeed. We hold every DraftKings and FanDuel slate back to 2016 — the game pools, the salaries, the projections, and the lineups the field submitted — and serve them through one API key over REST, GraphQL, MCP, or an SDK. Below is exactly what we have, sport by sport, including where it thins out.
Four layers of daily fantasy data.
Each layer keeps the operator's own values and joins to the same player and game ids the rest of the API uses.
Slates and game pools
49,496 DraftKings and FanDuel slates across NBA, NFL, MLB, NHL and PGA, each holding its game pool, scoring rules, and format — classic, showdown, single-game, late-swap. NBA runs from October 2016; the other four begin in 2021.
Priced player pools
10.7M pool entries carrying salary, position, roster position, team, injury status, and the operator's own scoring average. Loaded for NBA, NFL, MLB and NHL. PGA is slate-level only today.
Contests and submitted lineups
11,201 contests with entry fees, prize pools, and caps; 136,296 entrants; 6.8M submitted lineups; and 53.1M roster spots carrying rank, points, and winnings. Almost entirely NBA — 11,195 of the 11,201 contests.
Projections
2.5M projections carrying floor, ceiling, and standard deviation, each tagged with which of the 56 sources produced it. NBA only at present.
What we hold, sport by sport.
Published in full, gaps included. DraftKings and FanDuel are the two operators; no other site is covered.
The deepest sport by a wide margin. 24,586 slates from 1 October 2016 through the slates already posted for the season ahead, with 2.7M priced pool entries. Every contest, entrant, and submitted lineup we hold is NBA — 11,195 contests and 6.3M lineups — and so are all 2.5M projections.
GET /api/v1/dfs/slates?league=nbaGET /api/v1/dfs/contest_user_lineupsGET /api/v1/dfs/slate_player_projections14,465 slates from 31 July 2021, carrying the largest player pool we hold — 5.8M priced entries, more than any other sport, because baseball slates list deep hitter and pitcher pools. Contest, lineup, and projection data are not loaded for MLB.
GET /api/v1/dfs/slates?league=mlbGET /api/v1/dfs/slate_players5,169 slates from 12 August 2021 with 1.0M priced pool entries. Six NFL contests are loaded, carrying 464,349 submitted lineups — a real sample, but the contest layer is otherwise NBA. No projection data for NFL.
GET /api/v1/dfs/slates?league=nflGET /api/v1/dfs/slate_players4,664 slates from 9 October 2021 with 1.3M priced pool entries covering skaters and goalies. Contest, lineup, and projection data are not loaded for NHL.
GET /api/v1/dfs/slates?league=nhlGET /api/v1/dfs/slate_players612 slates from 5 August 2021, tournament by tournament. Slate-level only today — the priced player pools are not loaded, and there is no contest or projection data. The tournament, round, and shot data behind these slates is fully available under the PGA endpoints.
GET /api/v1/dfs/slates?league=golfGET /api/v1/pga/tournamentsHow the data comes back.
Plain REST over HTTPS, keyset pagination, JSON out. Slate, pool, and lineups are three endpoints; GraphQL returns the whole tree in one request.
API=https://api.stat-api.com/api/v1/dfs
AUTH="Authorization: Bearer YOUR_API_KEY"
# 1. every NBA slate on the night of Finals game 6
curl "$API/slates?league=nba&date=2026-06-13" -H "$AUTH"
# 2. the priced player pool for one of them
curl "$API/slate_players?slate_id=SLATE_ID" -H "$AUTH"
# → { "slate_players": [ {
# "display_name": "...", "position": "PG",
# "roster_position": "PG/G/UTIL", "salary": 9800,
# "avg_points": 44.6, "injury_status": null, ...
# } ], "next_from_id": ... }
# 3. the lineups entered into a contest on that slate
curl "$API/contest_user_lineups?contest_id=CONTEST_ID" -H "$AUTH"
The tools we provide.
DFS API reference →
The ten live dfs tables with runnable examples — slates, settings, game pools, player pools, projection sources, projections, contests, entrants, and lineups.
GraphQL →
Ask for the slate, its contests, and every lineup inside them in one request. Nested relations, keyset pagination, metered by rows resolved.
MCP server →
Point Claude, ChatGPT, or your own agent at the data directly. Ten tools over streamable HTTP; the catalog tools need no key.
SDKs →
Typed clients for TypeScript, Python, Go, Java, and .NET — pagination, retries, and errors handled, so a slate pull is one call.
Sheets →
Build a view in the browser, save it, and every sheet becomes its own endpoint. Useful when the question is faster to answer than to script.
Data dictionary →
Every dfs column, its type, and what it means — the reference you keep open while you write the query.
In development.
Everything above this line is available today. These four are being built on top of it.
First-party projections
Our own model, trained on the slate history, salary movement, and box-score data already in the API — alongside the 56 third-party sources we carry today, not instead of them.
State-of-the-art optimizer
Lineup construction against the live pool: salary cap, roster and position rules, exposure caps, stacking, and correlation between the players you pair.
Slate simulation
Monte Carlo over a full slate to produce outcome distributions per player and per lineup, so you can see the shape of a result instead of a single projected score.
Ownership projections
Forward-looking field ownership for an upcoming slate, trained on the 6.8M submitted lineups already in the dataset.
DFS data questions, answered.
Which operators and sports are covered?
DraftKings and FanDuel only — no other daily fantasy site. Slates cover NBA, NFL, MLB, NHL and PGA: 49,496 in total, with NBA from October 2016 and the other four from 2021. Priced player pools are loaded for NBA, NFL, MLB and NHL; PGA is slate-level only today.
Is this real ownership or projected ownership?
Real. It comes from 6.8M lineups that were actually submitted, broken into 53.1M roster spots carrying rank, points, and winnings. The contest and lineup layer is almost entirely NBA — 11,195 of the 11,201 contests, plus 6 NFL contests holding 464,349 lineups.
Do you provide projections?
Yes — 2.5M projections with floor, ceiling, and standard deviation, each tagged with which of 56 sources produced it. They cover NBA only at present. Our own first-party projections are in development, along with an optimizer, slate simulation, and forward-looking ownership projections. The data underneath stays queryable either way.
What is shipped versus what is coming?
Everything described above the roadmap section is queryable today: slates, priced player pools, contests, submitted lineups, and third-party projections. First-party projections, the optimizer, slate simulation, and ownership projections are in development and are marked as such — we do not list them as available, and we are not attaching dates to them.
Can I join DFS data to the rest of the API?
Yes. The same API key reads NBA possessions, NFL play-by-play, MLB Statcast, stadium weather, and betting lines. A slate player resolves to the same player id the league tables use, so scoring inputs and DFS pricing sit in one query.
How current is it?
Slates load ahead of lock, so upcoming slates and their priced pools are available before the games start. Historical slates stay queryable indefinitely.
What does it cost?
The free tier is 50,000 records a month with no card. Paid plans start at $49/mo for 5M records, and every plan includes every sport and every layer — DFS is not a separate add-on.
See the whole DFS stack.
50,000 records a month, free, no card. The DFS reference lists every table and column with live, runnable examples.