Network and request-header rules removed 277 of 372 browser-User-Agent requests (74.5%) from the Browsers category on the blog, providing clearer traffic insights for Cloudflare Worker analysis. Over two UTC days, 95 Browser HTML observations differed from 14 Cloudflare Web Analytics page loads. Network evidence caught 60 cloud-classified requests with navigation headers, exposing rule flaws now repaired.

Client identity requires distinct evidence; nine signature verifications identify signers, not readers using assistants. The 578-versus-52 daily client discrepancy initially alarmed the author, though daily identifiers don't equal visits and time windows differed. A September 2 incident revealed 100 of 113 daily clients loaded one page, while 156 of 164 browser-UA page observations lacked referrers.

Most bizarrely, one mobile client fetched 31 distinct pages in one timestamp second, prompting: 'i am seeing daily clients as 113 for today, and it seems unbelievable to me... like which articles are they reading, where are they coming from...'. Four decisions are needed for useful comparison: match UTC windows, compare page events with loads, report identifiers separately, and record exclusions. Cloudflare documents script blockers and caching as reasons beacons miss page loads.

A script can run in an automated browser with no universal ratio separating causes.