A plain guide to the numbers a fingerprint test shows you: what the hash means, what the entropy estimate claims, and which signals are worth worrying about.
On this page
A fingerprint result is not a score.
Running a browser fingerprint test is the easy part. Understanding what comes back is where most people stall, because a result page shows a strange alphanumeric string, a uniqueness estimate, and a long list of technical rows, with very little explanation of which numbers matter. This guide walks through each part of a result, and explains what you can and cannot conclude from it.
Read the hash first
At the top of most results sits a long string of letters and digits. The string is the fingerprint hash: a fixed-length value computed from all the individual signals your browser exposed.
Its job is to be an identifier, not a rating. A short or neat-looking hash is not better or worse than a long one, and a hash that changes is not necessarily an improvement — it may simply mean you switched browsers or installed a program.
The comparison test
The useful test is comparison, not inspection. Run the test twice on the same setup. If the hash is identical both times, your browser presents a stable identity, which is exactly what a tracker wants. If it changes between runs, some signal is being randomised, and a stable identifier is harder to build.
What the uniqueness estimate means
Alongside the hash you will usually see an estimate expressed in bits of entropy, or phrased as a ratio such as one in a large number.
Entropy is a measure of how many other people share your configuration. Each additional bit of it halves the crowd you belong to, so the figure grows quickly as more signals are combined. A low estimate means you blend into a large group, and a high one means your setup is rare.
Treat it as an order of magnitude
The figure is not a precise measurement. It depends on the population being compared against, which the test cannot know, so it is best understood as a rough signal of rarity rather than a hard count.
The individual signal rows
Below the summary you will find the ingredients. Each row is one thing your browser volunteered.
| Signal | What it describes | How much it changes |
|---|---|---|
| User agent | Browser, version and platform | Changes when you update |
| Screen and window size | Display geometry and window shape | Fixed unless you resize |
| Time zone and language | Locale settings | Rarely changes |
| Canvas hash | How your system renders pixels | Changes with drivers or fonts |
| WebGL renderer | Graphics hardware and driver | Changes with hardware |
| Font list | Installed typefaces | Changes when you install software |
| Audio hash | How your system processes sound | Very stable |
Some rows are descriptive and easy to understand, such as your user agent or your screen size. Others are technical values whose meaning is not obvious from the number, such as the canvas hash, which you cannot interpret by eye and can only compare between tests.
Ranking the signals by weight
Not every signal carries equal weight. A rough ranking by identifying power helps you decide where to spend effort.
- Rendering hashes. Canvas, WebGL and audio produce values that are highly specific to your machine and stable over time. These are usually the strongest rows.
- Font lists. The long tail of installed typefaces is a distinctive and persistent signal.
- Locale and time zone. Common settings narrow the crowd less, but an unusual combination of language and region stands out.
- Declared strings. User agent and platform are easy to read, but also easy to fake and commonly shared, so they matter less than their prominence suggests.
| Signal group | Weight | Why it matters | What changes it |
|---|---|---|---|
| Rendering hashes | High | Highly specific to your machine and stable over time | Drivers, fonts, hardware |
| Font list | High | Distinctive, and it persists | Installing software |
| Locale and time zone | Moderate | Common settings narrow the crowd less | Rarely anything |
| Declared strings | Low | Easy to fake and commonly shared | Updating your browser |
Signals that measure intent, not identity
Settings such as Do Not Track and Global Privacy Control are requests, not measurements, and they say more about your intent than your identity.
Reading a one in N claim honestly
Results that boast a ratio — one in a hundred thousand, one in a million — deserve a careful eye. The figure is a probability, not a promise.
It estimates how many visitors in a given population would share your exact combination, based on the signals measured and an assumed crowd. Change the assumed crowd and the ratio changes with it, even though your browser did not.
There is also a limit to what any single test can see. It measures the signals it knows how to measure, and a tracker on a real site may collect more. A modest ratio from a test is not a guarantee of anonymity, and an alarming one is not proof that you are uniquely exposed on every site you visit. Read the figure as an indicator of how distinctive your setup is, and nothing stronger.
Re-testing after you change something
The real value of a fingerprint test is before-and-after. Change one thing, re-run the test, and compare.
Change one variable
Reinstalling an operating system and a browser together tells you nothing about which one mattered.
Compare the hash first
If it did not move, nothing meaningful changed and you can stop there.
Watch the estimate, not just the rows
A row can shift without your uniqueness moving at all.
Expect surprises
Adding privacy extensions sometimes makes you rarer, not more common.
This loop is the only honest way to know whether a tweak helped. How to reduce your browser fingerprint covers the changes worth testing, and what is browser fingerprinting explains why some signals are inherently more stubborn than others.
The limits of a single test result
It is a measurement of one moment, on one page, against an assumed population.
| You can conclude | You cannot conclude |
|---|---|
| Whether your hash is stable across runs | Whether a real site will recognise you |
| Roughly how rare your setup is | Your exact place in a population |
| Which rows moved after a change | What other trackers collect |
Before you change anything
Work through this in order.
- Run the browser fingerprint test and note your hash and your uniqueness estimate.
- Run it a second time on the same setup and compare the hash.
- Change one thing only, then re-run and compare again.
- Open the canvas fingerprint test to see a single strong signal on its own, and hold that row against the summary.