Fingerprint entropy
A measure, in bits, of how much identifying information a browser attribute carries, and therefore how close a fingerprint comes to being unique.
Fingerprint entropy quantifies how identifying a browser attribute is, expressed in bits. Each bit halves the population that shares your value: an attribute carrying one bit splits users in two, while roughly 33 bits are enough to single out one person among everyone on earth. A common timezone might contribute a few bits; an unusual font list or a rare GPU string contributes many more. The EFF's Panopticlick project popularized this framing by measuring how much entropy ordinary browser configurations leak.
Entropy explains why fingerprinting defenses can backfire. Blocking canvas reads, running a rare browser, or installing unusual privacy extensions all make a configuration less common, which adds bits rather than removing them. The configurations that blend in are the boring, internally consistent ones that look like millions of stock installations. The same logic governs configured identities: a profile whose attributes are individually plausible but mutually contradictory is effectively a high-entropy beacon, because the contradiction itself is rare.
This is why Oculr derives every surface from a single profile definition: user agent, client hints, screen, GPU strings, timezone, locale, and geolocation all come from one schema, so they cannot drift apart between layers. Fingerprints are deterministic per profile, presenting the same identity every launch, and every release is regression-tested against public fingerprint test suites and real-device baselines.
