How to Score a Wrong Answers Only Game

Agree on self-voting and ties before collecting wrong answers. The scoring discussion explains the two-player problem and alternatives such as a host judgment or an unscored comparison, without inventing an outside audience to supply votes the room never cast.

Editorial credit: — Quirky Query Hub editorial pen name · Editorial & testing methods

Original illustration for How to Score a Wrong Answers Only Game

Four people submit four deliberately silly explanations. One receives two votes, and two others receive one each. The arithmetic is simple; the argument usually begins when somebody assumes that two votes should produce two points, a tie should require a winner, or an anonymous answer can never be linked to its writer. Explain those choices before the first submission and scoring becomes a small piece of administration rather than the main activity.

The current Wrong Answers Only tool gives one game point to each writer whose answer ties for the most votes in a completed round. Votes select the round's winner or winners; they are not directly added to the cumulative score. A writer with three votes and a writer who wins a different round with two votes each gain one point for their respective wins.

This guide separates that actual behavior from optional paper rules. It includes a complete fictional three-round ledger, a shuffled-answer mapping, and ways to handle misunderstandings without inventing a humor rating. The examples are fabricated demonstrations of the counting method, not evidence that one joke or scoring system is objectively better than another.

Decide what the vote means

Choose a criterion that ordinary players can apply. “Pick the answer you would like to hear expanded into a story” is usable because it asks for a preference. “Identify the scientifically funniest response” promises a measurement this activity does not provide. The browser counts selections; it does not analyze comedy, originality, writing quality, or audience reactions.

Use one criterion for the announced set of rounds. If the criterion changes, start a new clearly described activity. A host who secretly adds points for confidence or penalizes short answers changes the meaning of earlier results. That is especially confusing when somebody chose a quiet written response because performance was never part of the invitation.

State the self-voting arrangement. The current interface allows voters to select any submitted answer, including their own. It does not automatically enforce a no-self-voting rule. A group may agree not to self-vote, but that is a manual convention requiring cooperation. Present it honestly rather than claiming the button will disappear or an invalid choice will be rejected.

Follow the tool's two stages

The collection stage asks each configured player for a nonempty response in order. Choose Submit privately for that person's answer, then give the next person a chance to enter theirs. After all answers have been submitted, the tool switches to voting and shuffles their display order. Do not reveal a favorite early while the final writer is still thinking.

During voting, each configured player makes one selection in the displayed sequence. The tool stores the answer's author index behind the button, so a displayed label such as Answer 1 is not necessarily the first person who submitted. Read the actual answer text. Trying to infer the writer from the answer number defeats the purpose of a shuffled display and can produce an incorrect manual tally.

After the last vote, the tool counts votes per answer, finds the largest count, and adds one point to every writer at that count. It presents the winning player's name or the names of tied players and logs the result. Further voting does not add more points to the completed round. Use Next round when the group wants a new prompt, rather than attempting to reopen voting by repeatedly tapping an answer.

The workflow expects every configured player slot to submit and vote. A blank answer is not accepted as a built-in skip. People can still decline to participate; the host must pause and reconsider the arrangement instead of telling them the interface requires consent. Do not promise a withdrawal or edit control that has not been checked in the actual interface. An unfinished round can remain unfinished.

Read a shuffled set without losing its authors

Imagine four fictional players: Ada, Bruno, Ceri, and Dax. Their prompt asks why an imaginary staircase has opened a café. Ada writes “To offer a step-by-step breakfast.” Bruno writes “The landing needed somewhere to land.” Ceri writes “It serves coffee one level at a time.” Dax writes “The elevator wanted a competitor.” Each is one answer with one writer.

For this demonstration, the shuffled reveal order is Ceri, Ada, Dax, Bruno. Answer 1 therefore belongs to Ceri, Answer 2 to Ada, Answer 3 to Dax, and Answer 4 to Bruno. This is one possible ordering, not a guaranteed arrangement or a predictable shuffle pattern. A new round can place the same people in different positions.

Suppose voters select Answer 2, Answer 4, Answer 2, and Answer 1. The display-label counts are two for Answer 2, one for Answer 4, one for Answer 1, and zero for Answer 3. Translated to writers, Ada has two votes, Bruno one, Ceri one, and Dax zero. Ada receives one game point. Nobody receives a bonus for having been displayed first or last.

If a host reads answers aloud, read the labels consistently and avoid adding descriptions such as “this excellent one.” During a manual count, write both the answer label and its writer mapping before totals are transferred. The browser maintains that mapping internally; paper hosts need their own reliable record. Do not mix a previous round's label order into the current one.

A complete three-round ledger

All numbers in this example use four actual fictional voters, one vote each, and the tool's shared-win rule. Scores begin at Ada zero, Bruno zero, Ceri zero, and Dax zero. Round one is the staircase café round just described. Votes total four: Ada two, Bruno one, Ceri one, Dax zero. Its point awards are Ada one and everybody else zero.

Round two asks what a very formal potato puts on a business card. Ada receives two votes, Bruno two, Ceri zero, and Dax zero. Four votes have again been cast. Ada and Bruno both match the maximum of two, so each gains one point. The cumulative scores are now Ada two, Bruno one, Ceri zero, Dax zero. The tool does not halve their points because the round was tied.

Round three asks why an imaginary bridge has requested a day off. Each writer receives exactly one vote. The maximum is one, and all four writers match it. Each gains one point. The final scores are Ada three, Bruno two, Ceri one, and Dax one. A full four-way tie can award four points across the group even though the round is still just one round.

Check that ledger in two different ways. Each round's vote counts sum to four, because there were four voters. The point awards sum to one in round one, two in round two, and four in round three, because there were one, two, and four tied winners. Across the session, seven points are awarded. Final individual scores also sum to seven: three plus two plus one plus one.

That last check is important. Requiring total awarded points to equal the number of rounds would incorrectly reject a legitimate tied session. Conversely, directly transferring all twelve votes into cumulative scores would award the wrong quantity. Keep the words “votes this round” and “game points” distinct whenever reading results or writing a manual ledger.

Let a tie remain an outcome

A tie means the announced rule found multiple leaders. It does not mean the count is broken. The current tool shares the point among those leaders and moves on when the group selects a new round. It has no need for an imaginary outside voter, a host's extra ballot, or a secret judgment of which writer tried harder.

If a group wants a separate manual runoff, announce that before the session and distinguish its results from the browser score. The tool has already awarded its shared points. A later spoken preference does not automatically remove them or change the completed result. Running a paper tournament alongside a different on-screen tally can confuse players, so choose one primary record where possible.

With two players, self-voting can produce repeated one-to-one ties, and a no-self-voting agreement gives each answer one vote if both follow it. That is a predictable consequence of the format, not proof that either player failed to appreciate the other's humor. Two people may prefer an unscored comparison, or invite a willing third reader for a separately announced manual judging activity.

An all-way tie can be a satisfying finale. Read a detail from each answer that people choose to discuss, or combine compatible fictional ideas with their writers' agreement. Do not make the tied players compete again merely because the host planned to announce a single name. A shared result is already an understandable finish.

Understand anonymity before promising it

Anonymous display hides author names during the answer comparison and shuffles the order. It does not guarantee that writers cannot be recognized. A distinctive phrase, familiar reference, or self-identifying answer can reveal a person. On a shared device, other participants may also see information during handoffs if the group does not manage the screen carefully.

The result names the winning writer or tied writers even when answer display was anonymous. Do not promise that authors remain hidden forever or that names are revealed only after everyone gives a second approval. If continued anonymity is essential to the chosen activity, use a different manual arrangement whose records and reveal rules you can actually control.

Answers and results are part of local game state. “Private submission” is not a promise of deletion or secrecy from everyone with access to the device. Keep personal information out of the prompt and answer. Imaginary staircases, potatoes, and bridges supply enough material without identifying real people or quoting private communications.

Resolve mistakes without manufacturing a result

If a voter says they clicked the wrong answer, pause and inspect the supported controls. Do not promise an undo or silently assign a compensating vote to somebody else. If the completed result cannot be changed through a supported action, describe it accurately as the recorded result. The group can decide not to treat that practice round as part of a separately kept manual competition.

If someone disputes the arithmetic, recount selections where they are available rather than debating the quality of the answers. The logical questions are who received each vote, what the maximum was, and which writers matched it. “Everyone laughed at mine” is not evidence that a ballot was counted incorrectly. Laughter was never a second voting channel.

If the same writer wins several rounds, do not secretly change their points or give others consolation multipliers. Change the next activity openly: an unscored showcase, a new kind of prompt, or a cooperative miniature story. For prompt variety, see Categories for Unexpected Wrong Answers. Different prompts create different opportunities without rewriting the ledger.

If a response targets a real person or crosses an agreed boundary, pause before displaying it when possible. The scoring procedure cannot decide whether the material fits the room. Do not invent a software moderation guarantee. A host can stop the round, ask for a voluntary rewrite under a manual practice arrangement, or end the game. Recording a winner is less important than accurately describing what participants agreed to do.

Prepare a scoring briefing that fits in a few sentences

Try: “Each player submits one fictional wrong answer, then each votes once. The browser allows self-votes; our announced convention is the one we agreed before starting. Each answer tied for most votes earns its writer one point. Anonymous display hides names during comparison, but winning names appear in the result.” Follow with the criterion, such as choosing the idea people want to hear expanded.

Run one explicitly unscored practice exchange if the shuffled labels are unfamiliar. Ask someone to point out which text they would choose, without assuming its number identifies a writer. Explain that two votes produce one winning point, not two points. A small demonstration is more useful than discovering the distinction after a close final round.

Before beginning a session, check the participant roster, answer visibility choice, voting criterion, self-voting convention, tie treatment, and ending. During play, finish collecting before revealing and finish voting before interpreting the outcome. Afterward, use the recorded total without attaching claims about anybody's personality or talent. A preference count describes the selections made in that particular round.

For the wider event, read How to Host a Funny Answer Competition. For device handoffs, use How to Play Multiplayer Question Games on One Device. A well-run scoring system leaves a reader able to reconstruct the result, understand a tie, and return attention to the invented answers rather than argue about a rule announced too late.

Find your next good question

Your games, your device

Players, answers, favorites, and history stay in this browser. Export a backup before clearing browser storage.

Select a backup to preview its record count.

Recovery copies

Restore and clear operations keep a pre-change copy in this browser. Download one before restoring it.