Cricket Signals

Cricket signals expose collected social and news items together with derived match-level signals for cricket.

Backing views

signals.cricket_source_items
signals.cricket_match_signals

These views mirror the public.cricket_source_items and public.cricket_match_signals tables.

Suggested flow

1. Load cricket source items with the same match_id used by the main cricket feed.
2. Review recency, publisher, source, and extracted teams or players.
3. Load cricket match signals for the same match_id.
4. Rank signals by impact_score or confidence_score.
5. Join back through source_item_id when you need the original source item that produced a signal.

Supported source values

• x
• reddit
• google_news

Relationship between the views

• signals.cricket_match_signals.source_item_id points to signals.cricket_source_items.id.
• match_id is stored in both views so you can query either layer directly by match.
• Deleting a source item cascades to related match signals in the backing tables.

What the source items view contains

• ingestion metadata such as source, external_id, url, publisher, and published_at
• content fields such as title, content_snippet, author_name, and author_handle
• enrichment payloads such as engagement, teams, players, and raw_payload
• dedupe and audit fields such as dedupe_hash, collected_at, created_at, and updated_at

What the match signals view contains

• signal identity fields such as signal_type, entity_type, entity_id, and market_direction
• scoring fields such as sentiment_score, impact_score, and confidence_score
• human-readable explanation fields such as summary and reason
• additional enrichment fields such as tags, metadata, created_at, and updated_at

Notes

• All row ids are UUIDs generated with gen_random_uuid().
• JSON payload fields are returned as jsonb.
• Use the endpoint pages in this section for field-level filtering examples.