Answers comparison

Question by question: what you said, what they said, and where you split.

Decision: ⚪ idea

Design: Figma

Context

The other half of comparison mode, and the half that works in both modes - against a friend, and against an orientation. Every question the taker answered becomes a row that opens into the two answers beside each other.

How it behaves:

• A list of questions, collapsed - the statement is the row, and expanding it reveals the answers underneath.
• Two answers, one verdict - your answer and theirs sit side by side with a mark between them: a tick when they agree, a cross when they do not.
• The orientation is named - each expanded row says which orientation the question fed, so a disagreement is legible as a disagreement about something.
• Rows are tinted by outcome - agreement and disagreement are visible before anything is opened.
• Only disagreements - a toggle drops every row where the two of you matched, which is what most people are actually looking for.
• Search - a hundred questions is unbrowsable without it.
• No answer is a state - a question one side skipped shows as unanswered rather than as a disagreement.
• Multi-answer questions show every pick - each selected answer gets its own mark, so partial agreement stays visible.

Against an orientation the right-hand side is what that orientation implies rather than what a person said, which turns the same view into "where do I actually differ from progressivism".

Opportunity

• It is the evidence behind the result - every number on the result screen ultimately comes from these rows, and this is the only place a taker can audit it.
• Disagreement is the interesting part - the filter turns a long list into the three questions worth arguing about.
• It works without a second person - orientation mode gives a solo taker the same feature, which is most takers.
• It teaches what an axis means - seeing the questions behind progressivism explains the label better than a description could.
• Cheap to build on what exists - answers are already stored per question; this is a view over them.

Risk

• Answers are more sensitive than results - a score is an abstraction, an answer is a statement someone made, and this shows it verbatim to another person.
• It is the sharpest social surface we have - a filtered list of everything you disagree about is an argument generator, and we built the filter.
• Long quizzes make long lists - a hundred rows is a scrolling exercise, even with search.
• A tick reduces a nuanced answer - agreeing for opposite reasons still shows as agreement.
• Comparing to an orientation implies it has a position - what "progressivism would answer" is our modelling, presented as if it were a participant.