FERMATMIND · FRIENDSHIP QUIZ
Friendship Quiz Scores: Matches, Categories and Ties
Read your friendship quiz score, six-category radar, streak and per-question accuracy. Understand the average and tied ranks without rating your friendship.
Matches measure guessing, not similarity
The creator's own choices form the frozen answer key. A friend scores one point for each matching guess. If a friend prefers hot drinks but correctly guesses that you prefer cold drinks, that is still a match. Whether your own preferences are similar doesn't enter the score.
Percentages and playful titles
Percentage = matches ÷ questions × 100, rounded to a whole number. For example, 22/30 displays 73%, not the 73rd population percentile. 0/30 is 0%; 30/30 is 100%. Titles use that displayed percentage: 90%+ Detail detective, 73%+ On the same wavelength, 50%+ Getting the picture, 27%+ Plot twist explorer, and Fresh discoveries below that.
These are original game labels, with no psychometric validation. Long-time friends can miss answers: preferences change, a question may not fit, and someone's self-description can differ from what a friend sees.
How to read the radar and its comparison
The purple radar shape shows your match rate in each of six categories. A grey dashed shape appears only when other visible entries provide a comparison. Read each direction separately: a larger shape is not a relationship diagnosis. Each category uses its own matches divided by its question count. All categories tied for your highest rate appear as highlights. The comparison averages other visible completed entries on this exact frozen quiz, excluding you, hidden entries and creator previews. With no other entries, it says there is no average. This isn't a site-wide or population norm, and another device may create another entry.
Why guesses lock, and what the question breakdown means
Choosing an option commits that guess before the answer is revealed. Back lets you review, not rewrite it; a network retry cannot turn a miss into a match. Correct guesses build a streak and a miss resets it, while the result keeps your longest run. A refresh restores committed guesses for this browser key. The question breakdown uses all visible completed entries on this quiz, including yours, and lists the hardest questions first. Hidden entries and previews do not count; an empty sample gets no invented percentage.
Tied rankings and replaying
The friendboard sorts by matches and gives equal scores the same rank. Scores of 25, 25 and 20 appear as 1st, 1st and 3rd. Ordering within a tie doesn't show that one person knows the creator better. This browser retains its first completed entry; resubmitting the same choices returns that result. Another device or cleared storage can't reliably be recognized, so treat the ranking as a game, not a verified competition.
What to do with a surprising answer
Pick one question you both want to discuss: 'I guessed A because last time… What made you pick B?' Listen before reading a miss as a lack of care. The creator can make another round of the same length, prioritizing unused questions; longer rounds may reuse some. Or swap roles and let your friend guess you.
Everyone's choices and solo play
This answers a different question: which side creators chose for themselves, rather than how accurately friends guessed. It uses up to 200 recent active quizzes whose creators opted in, and shows a question only with at least 20 eligible answers. Counts are answers, not unique people. This small, self-selected sample doesn't represent a country, population or personality distribution. Insufficient samples stay empty; solo answers aren't uploaded.
FermatMind Editorial · Original game and guide · friendship-v1