CONTEXT UNKNOWN
Context Unknown
Everyone supplies a word without knowing the story — then the finished tale is read aloud
Players
4–10 players
Play Time
10min-
Everyone supplies a word without knowing the story — then the finished tale is read aloud
Create a room and send the code or the URL to your friends. One phone or computer each — play on a call or in the same room.
Join with a room code
Free, no sign-up. 4–10 players. The scripts are Japanese, so the game screen is in Japanese only. Browser-first — not in the app yet.
Context Unknown is a fill-in-the-blanks party game: everyone supplies single words without ever seeing the story, then the finished text is read aloud together. It ships with 58 scripts, and one is drawn at random each round. When a round starts, your screen shows only the blanks you are responsible for — their number and category, like “blank 3 / [animal]”. Nobody sees the script itself, not even its title. Each answer is one word, up to 10 characters, and ordinary words work best: trying to be funny is exactly what spoils it. Once everyone has answered, the reveal begins. The completed text appears character by character in time with the narration, like a quiz show reading out a question, and every device follows the host's narration so a remote group laughs at the same moment. The narration uses your browser's built-in speech synthesis, so there is nothing extra to install. After the reading, each player votes for the funniest blank (you cannot vote for your own). Votes are points, and only the results screen reveals who wrote which word. This is a browser-first title and is not in the dokopa app yet. No install: create a room and share the URL or room code. Note that the scripts are written in Japanese, so the game screen itself is Japanese only. For other games that build one piece of writing out of small contributions, dokopa also has Yomibito-Shirazu (a haiku passed around one character at a time) and Relay Novel (a story written one page at a time), both playable in your browser.
One script is drawn at random and its blanks are shared out among the players. All you see is the number of each blank you own and its category — never the script, not even the title. Answers are one word, up to 10 characters (60 seconds by default; blanks left empty are filled automatically, so nobody has to wait for a missing player).
The finished text appears character by character in time with the narration — the title first, then the body. Words that filled the blanks are highlighted. Every screen follows the host device's narration, so a group on a call laughs at the same line. Each player can turn the narration on or off and pick the volume and voice.
Once the whole text is out, everyone casts one vote for the blank that made them laugh most. You cannot vote for a blank you filled yourself, and authors are still secret at this point.
The votes per blank are shown, and only now is it revealed who wrote each word. Votes are points, and a running total is kept across rounds. The next round draws a different script.
Votes cast for the words you wrote are your points — highest total across the rounds wins
Don't strain for a funny word. Plain ones — “umbrella”, “elbow”, “bathhouse” — are the ones that transform once they land in a sentence.
Categories are loose ([animal], [food], [body part]). When in doubt, submit the first thing that comes to mind.
Playing in the same room? Turn the narration on for just one device so it stays intelligible (it is on by default and each player can toggle it).
On a video call, leaving it on for everyone is fine: every screen follows the host device's narration, so nobody runs ahead.
Three to five rounds works best — it evens out the luck of the draw, and the blanks are reassigned every round.
Filling in blanks is an old idea, but the best moment is concentrated in one place: reading the finished text out loud. In practice that is exactly where these games get awkward — who reads, the reader sees the text first, and nobody enjoys performing. Context Unknown hands the reading to the machine.
During the reveal the finished text appears character by character in time with the narration, starting from the title, the way a quiz show reads out a question. The words that filled the blanks are highlighted, so you can see at a glance where somebody's handiwork landed.
Every screen follows the host device's narration, which means a group spread across a call laughs at the same word at the same time. The narration uses the speech synthesis already built into your browser, so there is nothing extra to install. Each player can toggle it and choose the volume and voice, and on browsers without it the game still works with the text alone.
Two-hander scripts are read with a different voice per speaker. The name is shown only on that speaker's first line; after that the voice and colour tell you who is talking.
Context Unknown is a browser-first release. No install needed — create a room on this page and start playing.
It is not in the Dokopa app yet.
