Data sources

Every game page on this site is built from public data. Here's the complete list of sources the pipeline reads, what each one provides, and which sections use it. For how these sources are processed, see the methodology page.

Critic scores & reviews

Source Provides Used in
OpenCritic Aggregate critic score, tier (Mighty/Strong/Fair/Weak), top review excerpts Verdict
RAWG Metacritic score, community rating, genres, tags, platforms, release date, playtime estimate, PC requirements Verdict, Similar Games, Stats Sidebar

Community discussion

Source Provides Used in
Exa.ai (semantic search) Recent Reddit threads, blog posts, and forum posts about each game — sourced via semantic search rather than direct API Community Pulse, Guides & FAQ, What they're asking (Reddit leg)
xAI Grok (x_search tool) Recent X (Twitter) posts discussing each game — retrieved via Grok's Agent Tools API with paragraph-level citations back to the original posts What they're asking (X/Twitter leg)

Live stats

Source Provides Used in
Steam Web API Current concurrent player count Stats Sidebar (live count + 7-day sparkline)
ProtonDB Linux/Steam Deck compatibility tier (Platinum, Gold, Silver, Bronze, Borked) Stats Sidebar

News & patch notes

Source Provides Used in
Steam News RSS Official developer patch notes, event announcements, and update posts What's Happening Now
Gaming news RSS (IGN, PC Gamer, Kotaku, Eurogamer, GameSpot) Publication coverage of major game updates, controversies, and announcements What's Happening Now

Streaming & trending

Source Provides Used in
Twitch API Live viewer counts, active stream counts per game Candidate ranking (game selection)
YouTube Data API Top video metadata for trending game content Candidate ranking

A note on Reddit

Reddit blocks direct API calls from Cloudflare Workers (HTTP 403). Rather than scraping or using workarounds, this site accesses Reddit content indirectly through Exa.ai's semantic search, which indexes public Reddit threads. This means community sentiment data is sourced through a search layer rather than pulled from Reddit's API directly.

A note on X (Twitter)

This site does not access X directly. Instead, xAI's Grok model runs the x_search tool against X's corpus and returns a summary with citation links to the original posts. Every X citation on this site links back to its source on x.com — author handle and post preserved — and each cited post retains its original publication timestamp (decoded from the X status ID). If a linked post is later deleted or made private, the link will 404 but the site's cached snippet remains until the next pipeline refresh.

Missing something?

If you know of a public data source that would improve game pages — price trackers, compatibility databases, community hubs — email [email protected].