Controlled-visit evidence · 6 canonical sites
What dating sites loaded
Descriptive results from one newest eligible passive visit per site, not a ranking, privacy grade, causal claim or representative sample of all dating sites.
Newest included observation:
Observed medians
Median across this curated sample.
Medians summarize the included site-level observations. They do not describe every visit to these sites or the wider category.
- Third-party requests
- 48
- Third-party tracking-service requests
- 17
- Third-party cookies
- 56 of 6 sites had complete cookie evidence
Included evidence
6 site-level observations
Sort any column. Open a profile for the currently retained, versioned report history.
6 sites
| bumble.combumble.com recorded 40 fewer third-party requests in the visit configured for Brave-list blocking.3 reports · Open latest evidence | 40 | 29 | 9 | Aug 17, 2026Brave-list blocking comparison · desktop |
|---|---|---|---|---|
| eharmony.comeharmony.com recorded 48 fewer third-party requests in the visit configured for Brave-list blocking.3 reports · Open latest evidence | 56 | 20 | 8 | Aug 17, 2026Brave-list blocking comparison · desktop |
| grindr.comgrindr.com recorded 16 fewer third-party requests in the visit configured for Brave-list blocking.3 reports · Open latest evidence | 68 | 4 | 0 | Aug 17, 2026Brave-list blocking comparison · desktop |
| hinge.cohinge.co recorded 56 fewer third-party requests in the visit configured for Brave-list blocking.3 reports · Open latest evidence | 58 | 18 | 6 | Aug 17, 2026Brave-list blocking comparison · desktop |
| pof.compof.com recorded 33 fewer third-party requests in the visit configured for Brave-list blocking.3 reports · Open latest evidence | 33 | 15 | 3 | Aug 17, 2026Brave-list blocking comparison · desktop |
| tinder.comtinder.com triggered fingerprint-like browser API patterns.3 reports · Open latest evidence | 2 | 0 | 0 | Aug 17, 2026Brave-list blocking comparison · desktop |
How to read this page
Eligibility and limits
- Each row is the newest successful, request-complete, uncapped passive lead visit for one canonical site.
- Every row here was measured under one cohort (shields-request-context-v2-adblock-rust-0.13.2-request-method-v1-playwright-1.62.1+subject-validity-v2+detector-coverage-v2, schema v1, producer unrecorded, GPC not requested), which is this page's denominator. Another category can publish a different cohort, so these medians are comparable within this page and not against another category's.
Full measurement identity
shields-request-context-v2-adblock-rust-0.13.2-request-method-v1-playwright-1.62.1+subject-validity-v2+detector-coverage-v2, schema v1, producer unrecorded, GPC not requested, legacy catalog-metadata hash 5a5b8699d87581ae2f9b4a98d622df2b6c78a9259fad43d922b1cb6eda677526, ServiceRole taxonomy service-role-taxonomy-v1 digest dfccf71d4119c154e71bf7908dd2914557e8fc981951941594b16b00b712ed67, metric contract metric-contract-v1 digest 19bd5b62d1530480481080a1f208d2664e3769ac16300560a1f3df5894eb9680
- Failed loads, incomplete recordings and accept/reject consent-interaction arms are excluded.
- Third-party tracking-service request counts are lower bounds derived from retained request rows whose recorded host matched a reviewed service-catalog suffix and from the recorded ServiceRole taxonomy.
- Category membership is editorial. The corpus is curated and is not a random or representative web sample.
- A controlled visit is one observation. Ads, experiments, caching, location and bot detection can change what loads.
- Blocking pairs simulate Brave's ad-block engine and default lists in the scanner; they are not live Brave-browser visits or counts of individually blocked requests.