With a blocking list

The same pages, loaded again while blocking the requests that a public privacy filter list targets. It shows how much of the third-party traffic such lists are designed to stop, and what still loads. It is not a recommendation of any product.

How this test works, and what it does not say

• We download a public privacy filter list when the scan runs (it is not stored in our repository) and apply only its simplest rules: whole domains, optionally limited to third-party requests or to some resource types. Rules with paths, wildcards, page elements or site-specific conditions are skipped. A real blocker using the full list would usually block more.
• Blocked requests are recorded and aborted. “With” counts what still went out. Tracking services are still counted with our own classification list, which is independent from the filter list.
• The two measurements are taken at different moments (up to a week apart), so ordinary run-to-run variation is part of the difference. Each figure is a median of several passes.
• Blocking can change how a page works (video, comments, paywalls). We do not measure that.
• Many browsers and extensions use lists like this one. Naming the list is not an endorsement of any product, and we receive nothing from any of them.
• This test is separate from the weekly ranking and never changes it. Raw reports: data/protected/<origin>/<date>/; table: protection.csv.