How your privacy score is calculated
The 0–100 privacy score on How Private Am I? is an inference-weighted estimate of how identifiable your browser looks to a typical website — not a security audit or malware scan. Higher is better. This page explains the formula and lets you model improvements.
Scoring pipeline
- Inferences — We turn User-Agent, HTTP headers, JavaScript signals, and optional IP geolocation into attributed findings (each with a confidence level).
- Exposure points — Scored inferences add points; protective inferences (GPC, blocked probes, etc.) subtract. Fingerprint entropy and core probe coverage add exposure.
- Consistency & profile credits — Mismatches between signals can reduce exposure; privacy-hardened browser families (Tor, Firefox RFP, Brave) receive profile credits.
- Net exposure —
net = max(0, exposure − protection − credits). This raw number is shown as exposure points in the API. - Calibrated score — Net is mapped to 0–100 with a curve so everyday browsers land in a meaningful mid-range while high-risk profiles still score low.
Exposure bands: high (score ≤ 40), moderate (41–65), low (> 65). See also anti-fingerprinting, GPC, and Tor Browser guides for practical fixes.
Inference roles
| Role | Effect on score | Examples |
|---|---|---|
| Scored | Increases exposure | Canvas fingerprint, WebGL renderer, IP country, Client Hints |
| Protective | Decreases exposure | GPC, DNT, blocked fingerprint probes, WebGL privacy mode |
| Indirect | Drives consistency credits | UA/Client-Hints mismatch, timezone vs IP |
| Display only | No points | Rendering engine label, compression support |
Interactive score simulator
Toggle improvement actions below to see estimated score changes for a typical Chrome desktop profile. Open this page from your analysis results to start from your live score. Re-run analysis after changing settings — estimates assume full implementation.
Limitations
Is this a security grade?
No. The score reflects signal exposure and fingerprint strength narratives on this site. It does not scan for malware, password reuse, or account breaches.
Why did my score change after an update?
We recalibrate the display scale periodically so typical browsers are not clustered near zero. Raw exposure points remain comparable; the 0–100 mapping may shift.
Do trackers use the same weights?
Unlikely. Real ad-tech stacks combine signals differently. Use this score for self-comparison and education, not as a guarantee of anonymity.
Measure your real browser signals.
Run live analysis