CS2ED outplayed? · misplayed? · CS2'd? Public methodology
Product methodology

What a report is allowed to mean.

CS2ED is the match-analysis product plane. It consumes verified identities and published parser artifacts. It does not decode .dem bytes in the browser, and it does not invent a second store of match truth.

Complete-chain claim refused.

The CS2 evidence vertical is blocked at the parser. SUPPORTED_BUILDS is empty in cs2-demo-parser 8fcc710bde2b72bfff919f73233a392c095e769b. SteamCloud steamcloud.demo.acquire.v1 is BLOCKED/NOT QUALIFIED. This site will not treat a fixture, labeled demonstration, or canary envelope as a verified match report.

A match is one fact

A ten-player Premier match is one n-ary participation fact, not a mesh of “played with” pairs. Durable rows live in steam-hypergraph PostgreSQL 18. CS2ED may project a report after SteamGraph (target) or Steam Hypergraph (current compatibility wire) publishes an admitted artifact. Unpublished, unknown, or unverified artifacts stay off this site.

How a match becomes a report

  1. Identity. Steam authenticates. SCDB owns the canonical subject. CS2ED has no /auth route in this release.
  2. Source authorization. The cs2ed extension retrieves Premier match codes and, separately, a user Faceit API key. This page never collects those secrets.
  3. Acquisition. Target SteamCloud steamcloud.demo.acquire.v1 would return an object digest. That action is not qualified. CS2ED does not GET demos on this Worker.
  4. Parse. Parser products emit immutable artifacts. CS2ED does not spawn a parser and does not parse in the browser.
  5. Verification and publication. Artifact admission and semantic publication are separate decisions. Unknown builds stay unpublished.
  6. Projection. Only after a confirmed publication may this product compose a report. V1 canary retrieval checks run/source and non-empty revision fields on the private origin, and only when ORIGIN is bound.

Diagnosis labels

OUTPLAYED, MISPLAYED, CS2D_REVIEW, and UNCLASSIFIED are CS2ED product labels. They are not parser output.

  • Outplayed? Evidence supports an opponent-created advantage and no stronger player-correctable cause. Next action: adapt.
  • Misplayed? Evidence supports a repeatable correction. Next action: fix.
  • CS2'd? Low-control or ambiguous moment. This is not a CS2 bug, desync, cheat, hit-registration, or Valve verdict. Tick-level research belongs to Subticked.
  • Unclassified. Dependencies or confidence are insufficient. Missing evidence cannot become a zero or a stronger claim.

Every metric on a published report must name its denominator, scope, and source. Correlations are not causes. Parser gaps are gaps: this product does not interpolate unobserved coordinates as facts.

Demonstration versus publication

The dashboard, report, maps, and trends shells may show a contained, labeled demonstration so the product language is reviewable. Those numbers are not live player data and are not a matches-analyzed total. Canonical report URLs use /report/:reportId. Unknown IDs are true 404s on this Worker; fixtures are not substituted. The labeled demonstration is only /report, /report/demo, and /report/demo-mirage-13-10.

Also read data status and privacy.