Alok Tomar
All case studies

Founder

SpreadModel

Sports projections API for the NFL, NBA and college football, graded in public against the closing line.

The product

SpreadModel sells projected scores, spreads, totals and win probabilities for every NFL, NBA and college football game to businesses through one REST API. Projections are re-run after the injury report, published before kickoff, never edited, and graded in public against the closing line.

What I built

  • Projected home and away points, margin, total, the uncertainty of each and the home win probability for every game
  • Re-run after the injury report: NBA every game day, NFL after the final Friday reports, college football every week
  • Published before kickoff and never edited; the last version before kickoff is the one that gets graded
  • Walk-forward backtest: average margin error within 0.15 points of the closing line for the NFL and 0.17 for the NBA
  • REST endpoints for projections, every version of a game, graded history and the track record, with API keys, rate-limit headers and usage counts

Try it here

Demos of tools from SpreadModel

Rebuilt with mock data so you can use them without an account. Fictional players, simulated prices.

Sports data API

$ curl "https://api.example-sportsdata.dev/v1/odds?game_id=g2" \
  -H "Authorization: Bearer sk_demo_••••"
200 OK· 38 ms · x-ratelimit-remaining 599
{
  "game_id": "g2",
  "matchup": "BUF @ KC",
  "spread_home": -2.5,
  "total": 48.5,
  "no_vig_home_win": 0.573,
  "books": 8
}
Open the demo : Sports data API