Odds Data API for Sports Betting Models and Backtesting

Features, labels and a closing-price benchmark, joined by shared ids.

A betting model needs features (prices), labels (outcomes) and a benchmark to judge itself against (the closing price). SockOdds v1 provides all three under the same ids, so building a training set is a join rather than a matching project. Price history and opening and closing snapshots are on the Pro plan, and history goes back to 22 December 2025 for some leagues (dates by league).

Why SockOdds

Prices as features. Every bookmaker's decimal_odds and implied_probability per outcome, plus line for handicaps and totals.
Movement as features. GET /odds/history returns every recorded change for a game, so opening price, size of move and time before the start are all things you can derive.
Labels graded from official sources. GET /results gives each outcome a verdict of won, lost, void, dead_heat, half_won or half_lost. Outcomes the evidence cannot settle are left out, not guessed.
A closing benchmark. snapshots=closing gives the last price recorded before the start for each quote, the usual yardstick for a model's prices.
The same ids everywhere. Event, market and outcome ids match across odds, history and results, and across bookmakers. The JSON loads into any data frame library.

What you can build

A training set

Page through /events for a date window, pull each game's history and results, and join on outcome id. Treat an outcome with no verdict as unlabelled, never as a loss.

A backtest

Replay your model's picks against the prices recorded at the time, settle them with /results, and report both return and closing line value.

A calibration check

Compare your predicted probabilities with the closing implied_probability once the margin is removed, and with what actually happened.

Try it

This request asks for Sportsbet's closing winner prices on a finished NFL game.

curl -sS "https://api.sockodds.com/v1/events/ev_000ckk0h3n6r/odds?markets=winner&bookmakers=sportsbet&snapshots=closing" \
  -H "Authorization: Bearer $SOCKODDS_API_KEY"

const response = await fetch("https://api.sockodds.com/v1/events/ev_000ckk0h3n6r/odds?markets=winner&bookmakers=sportsbet&snapshots=closing", {
  headers: { Authorization: `Bearer ${process.env.SOCKODDS_API_KEY}` },
});
console.log(JSON.stringify(await response.json(), null, 2));

import json, os, urllib.request

request = urllib.request.Request(
    "https://api.sockodds.com/v1/events/ev_000ckk0h3n6r/odds?markets=winner&bookmakers=sportsbet&snapshots=closing",
    headers={"Authorization": "Bearer " + os.environ["SOCKODDS_API_KEY"]},
)
with urllib.request.urlopen(request) as response:
    print(json.dumps(json.load(response), indent=2))

Response from a local copy of the API with invented test prices

{
  "data": {
    "id": "ev_000ckk0h3n6r",
    "name": "Arizona Cardinals v Atlanta Falcons",
    "league": {
      "key": "nfl",
      "name": "NFL"
    },
    "scheduled_start": "2026-10-12T08:00:00.000Z",
    "status": "final",
    "participants": [
      {
        "role": "away",
        "identity": {
          "key": "atlanta_falcons",
          "name": "Atlanta Falcons"
        }
      },
      {
        "role": "home",
        "identity": {
          "key": "arizona_cardinals",
          "name": "Arizona Cardinals"
        }
      }
    ],
    "markets": [
      {
        "id": "mk_s4fv79z47rbkwev7",
        "key": "winner",
        "name": "Winner",
        "period": "full",
        "subject": null,
        "alternate": false,
        "outcomes": [
          {
            "id": "oc_53f4yzaqeyex1xhg",
            "name": "Arizona Cardinals",
            "side": "home",
            "line": null,
            "prices": [
              {
                "bookmaker": "sportsbet",
                "decimal_odds": null,
                "implied_probability": null,
                "available": false,
                "closing": {
                  "decimal_odds": 1.72,
                  "implied_probability": 0.581395,
                  "observed_at": "2026-10-12T06:00:00.000Z"
                },
                "observed_at": "2026-10-12T06:00:00.000Z"
              }
            ]
          },
          {
            "id": "oc_1xh21t2rzp7ss665",
            "name": "Atlanta Falcons",
            "side": "away",
            "line": null,
            "prices": [
              {
                "bookmaker": "sportsbet",
                "decimal_odds": null,
                "implied_probability": null,
                "available": false,
                "closing": {
                  "decimal_odds": 2.12,
                  "implied_probability": 0.471698,
                  "observed_at": "2026-10-12T06:00:00.000Z"
                },
                "observed_at": "2026-10-12T06:00:00.000Z"
              }
            ]
          }
        ]
      }
    ]
  },
  "meta": {
    "count": 1,
    "cursor": null
  },
  "links": {
    "next": null
  }
}

Each outcome has available: false and a null current price, because the game is over, and a closing object with the last price recorded before the start. Join these outcome ids to /results?event_id= for the labels, and to /odds/history for the movement that led there.