Two addresses can look almost interchangeable while supporting different research conclusions. Imagine an analyst finds a familiar brand in an official document, then labels a similarly named address the best baccarat site. The error occurs before any discussion of quality: the evidence has been attached to the wrong object.
Follow the printed identifier#
A hostname is an identifier, not decorative punctuation. Preserve the spelling, suffix and subdomain shown by the source. Separately record the address being discussed. An apparent relationship between them becomes a question to resolve rather than an equivalence to assume.
A concrete example comes from the October 29 GSA list, retrieved on 11 November 2026. The preserved research records OKBet entries including www.okbet.com. That is not evidence, by itself, that the bare hostname has been independently verified or that an observed redirect is authorized. No redirect chain was tested for this article.
The point is not that one form is necessarily unsafe. It is that the source and conclusion must describe the same thing. A comparison loses auditability when it quietly normalizes away distinctions that the original document preserved.
A three-column correction#
Use a worksheet with “printed hostname,” “claimed hostname,” and “relationship evidence.” If the first two differ, leave the third unresolved unless an authoritative record explains the relationship. A working page or a visually identical design does not establish regulatory continuity. Even a genuine technical redirect would show routing, not automatically the full scope of authorization.
Keep brand spelling in another field. A brand may have multiple listed addresses; an address can also appear within a complicated operator row. Neither fact allows a publisher to extend the entry to every plausible brand-related string.
What changes in the finished comparison?#
Replace “the brand is verified” with a bounded statement about the exact record. Withhold claims about other hostnames rather than labeling them illegal. This preserves the distinction between an unsupported association and a demonstrated adverse finding.
These notes remain local. The 24 June 2026 ASC circular, retrieved in the 11 November corpus, prevents assuming that game-focused comparison copy is publishable advertising. Nor would a precise hostname establish eligibility: PAGCOR’s 2021 code, retrieved the same day, excludes students at Philippine educational institutions and people under 21. Identity precision solves one research problem, not every regulatory or audience question.