§4 · evil rating telemetry

telemetry.evilbyte.net

Every 15 minutes this pulls new HTTP request logs from Cloudflare Log Explorer across the account, computes an Evil Rating for each distinct requestor with the same formula the calculator uses, and keeps one latest-seen row per client IP. This page is what that data actually looks like — not a demo, not a synthetic vector.

Loading…

Rating distribution

Every distinct requestor currently on file, bucketed by §3.3's bands.

Loading…

Does the formula track Cloudflare's own signals?

Average Bot Management score and WAF attack score, grouped by the same bands above. Neither one is a formula input — §4.5.2's "Presumption of Evil" stands unmodified — they're recorded purely for comparison. If the Evil Rating means anything, the Evil/Saturated Evil bands should skew toward Cloudflare's own bot/attack scores too.

Bot Management score

Loading…

WAF attack score

Loading…

Top Autonomous Systems

By distinct-requestor count. F_AS comes from the ERA, not from this table.

Loading…

By zone

By distinct-requestor count, across every zone on the account.

Loading…

Methodology & known gaps

This pipeline reconstructs the six factors from historical log rows, not from a live request — a few of them are necessarily approximations of what the live MITM would compute in the moment.