@buzzr/entertainment-engine

Transparent sports entertainment scoring and hybrid ML prediction utilities for Buzzr.

The package is intentionally pure: it does not import React Native, Expo, Supabase, AsyncStorage, or app services. Apps and jobs provide data through plain objects, then persist results however they choose.

Core API

import {
  resolveBuzzScores,
  enrichGameRowWithBuzzScores,
  predictGameWithDiagnostics,
} from '@buzzr/entertainment-engine';

const resolved = resolveBuzzScores({
  league: 'NBA',
  status: 'scheduled',
  startsAt: '2026-06-12T01:00:00Z',
  homeTeam: 'Boston Celtics',
  awayTeam: 'Los Angeles Lakers',
}, { upcomingLike: true });

const prediction = predictGameWithDiagnostics({
  league: 'NBA',
  status: 'scheduled',
  startsAt: '2026-06-12T01:00:00Z',
  homeTeam: 'Boston Celtics',
  awayTeam: 'Los Angeles Lakers',
}, {
  odds: { spread: -1.5, overUnder: 226.5 },
  teamPower: { home: 9.2, away: 8.7 },
});

Design

• resolveBuzzScores is deterministic and explainable.
• predictGameWithDiagnostics layers trained weights over transparent features and returns factor impacts, input coverage, confidence, and model version metadata.
• Training helpers accept already-built examples; Supabase extraction belongs in the consuming app or job.