Name-Seeded Result Generator

What this generator does

Normalises whatever you typed (trims it, collapses extra spaces, lower-cases it), then hashes it twice with two different salts to independently choose a primary entry — from a bank where some entries are deliberately rarer than others — and a secondary entry from a separate bank, and combines the two into one result. The same normalised input always produces the same two hash values, and the same two hash values always pick the same entries.

How to use this tool

1. Choose a theme.
2. Type a name or short phrase — your own, a character's, anything.
3. Press the button to see the result: a primary result and a secondary trait, each derived from what you typed.
4. Type the same name again, or send someone else the name to type, and check the result matches.

Understanding the controls

Theme

Each theme has its own two content banks (a primary, weighted by rarity, and a secondary, drawn evenly), combining into several hundred distinct results.

Name or phrase

Up to 80 characters. This is the seed — there is no separate seed field, because the name is the seed.

Common use cases

• A shareable, repeatable result for a name, unlike this site's other random pickers
• A quick icebreaker where everyone types their own name and compares
• Testing that two different capitalisations of the same name land on the same result
• A stable result to reference again later without needing to save anything

How this generator works

What this needs to be true is not that one result looks right, but that the whole mapping from names to results behaves properly, so that is what was checked rather than assumed. Determinism is checked directly: the same name, hashed any number of times, produces byte-identical output. Fairness is checked statistically, not by eye — the primary bank's rarity tiers are drawn at declared relative weights (5:2:1 for common:uncommon:rare here), and running the draw over 4,000 distinct sample names lands each tier's observed share within a few percentage points of what its weight predicts, for both themes. The primary and secondary picks are salted differently precisely so they do not move together: which primary entry a name draws says nothing about which secondary entry it will also draw, checked by confirming that names sharing a primary entry do not all share the same secondary entry too.

Randomness and fairness

Nothing here is random in the usual sense — a name is not a seed you supply alongside genuine randomness, it *is* the entire input. The mapping uses FNV-1a hashing with an extra mixing round, not a cryptographic hash: it is built for even distribution across names, not for resisting someone deliberately searching for a name that produces a chosen result.

For how randomness is produced across the whole site, see how Generate Random works.

Limitations and good to know

• Two different names can land on the exact same result. Each theme offers several hundred distinct combinations, but with enough different names tried, a genuine collision will eventually turn up — that is an expected property of any fixed-size mapping, not a defect.
• The mapping is not cryptographically secure. It is unsuitable for anything where it matters that a result cannot be deliberately searched for or reverse-engineered.
• Only two themes are offered, and neither can be extended by the person using the page — the content banks are fixed.
• Punctuation and non-alphabetic characters are hashed exactly as typed and are not stripped, so "Sam" and "Sam!" are treated as different inputs.

Privacy and your data

The name you type is hashed and matched against the content banks entirely in your browser. It is never transmitted, logged or stored.