← Back to Homepage
Changelog - Record of Technological Evolution

System Updates & Changelog

The roadmap and update history for our analysis pipeline, publication checks, database, and UI.

Current Total Analyzed Titles in Database
Unavailable
v2.3.02026-07-20

Review Evidence and Publication Hardening

Aligned public documentation with implementation and blocked analyses whose evidence cannot be verified.

  • Limited AI input to 40 representative reviews after duplicate and short-text removal
  • Bound overall reasoning, pros, and cons to review IDs and exact source quotes
  • Limited Metacritic / OpenCritic comparisons to successful retrievals
  • Removed unsupported forecast fields from public results
v2.2.02026-06-02

GEO/SEO Breakout Update (Phase 2)

Semantic optimization and evidence layer enhancements targeting AI search agents (AI Overview, Perplexity, ChatGPT).

  • Extended Dataset schema with variableMeasured property for statistical transparency
  • Added metrics explanation pages (/trust-score, /fraud-score, /evaluation-gap) with dynamic FAQPage JSON-LD schemas
  • Integrated dynamic macro statistics (total games, average fraud, average regional gap) in the Monthly Report
  • Added automated FAQPage schema embedding on Versus comparison pages
  • Opened Dataset JSON export API and dedicated data center page (/exports)
  • Created developer-focused API documentation page (/api/docs)
v2.1.02026-04-12

Edge Caching & Performance Optimization

Refined caching, Sitemap generation, and language routing for the Edge Runtime.

  • Optimized CDN cache-control headers for Cloudflare Pages
  • Implemented memory-safe Sitemap generator to prevent build crashes
  • Improved multi-lingual subpath routing in Edge Middleware
v2.0.02026-02-07

Introduced Review-Grounded AI Analysis

Structured representative reviews for AI processing and based analysis output on the supplied review text.

  • Processed representative samples containing Japanese and English reviews
  • Added conditional comparison of successfully retrieved Metacritic / OpenCritic scores
  • Added six-axis quality estimates based on review text
v1.0.02026-01-15

Game Evaluation Lab Launch

Initial public launch of our AI-powered Steam review integrity platform.

  • Implemented rule-based fraud scoring and six-axis quality estimates
  • Displayed a list of recent publishable analyses
  • Native 4-language support (ja, en, ko, zh)