search the curve
Add a word to the count.
Try a word, see whether it is new or rediscovered, and add one anonymous signal to the shared ranking.
how it works
Random searches become a ranked public signal.
Every valid search checks one dictionary boundary. A first match creates a public word record; later matches increment that word’s rediscovery count.
Nothing is weighted or predicted. The curve is made from the order people actually search: familiar words pile up, while rare rediscoveries stretch into the tail.
live rank curve
The experiment is the graph.
The left edge tracks the most rediscovered words. The long right side is where one-off searches wait to be found again.
Zipf trace
Rediscovery counts by rank
rediscovery leaders
Most searched words
- apple4
- discord2
- hello2
- water1
- tree1
- world1
- Jeffery1
- you1
discovery archive
New first finds keep the archive growing.
boundaries
Searches stay public, plain, and dictionary-bound.
Valid searches
Latin-script words with optional internal apostrophes or hyphens, present in the imported dictionary data.
Not counted
Numbers, emoji, punctuation-only input, overly long strings, spaces, and words absent from the current corpus.
Public notes
Anonymous notes are immediate, length-limited, plain text, and intended for word context rather than private messages.