What every metric and chart means, in plain language.
Your mastery adjusted for how recently you practiced. It decays by 0.5 points per day of inactivity (max 25 points). Practice regularly to keep it accurate.
Raw mastery before time decay. Starts at 50 per subtopic. Correct +6, skip −2, incorrect −8.
How long you took vs the question’s expected time. Below 1.0 is faster; above 1.0 is slower. Used to calibrate difficulty.
Active days in the last 14: Sporadic (0–2), Building (3–5), Steady (6+).
Change in accuracy over the last 7 days vs the previous 7. Positive means improving.
Share of subtopics you have practiced at least once. Higher coverage means fewer blind spots.
Subtopics practiced before but not recently enough; effective scores are decaying.
When a question returns for review: incorrect in 12h, skipped in 8h, correct in 3–14 days.
Projected mastery decay over 14 days without practice. The weak threshold is around 40%.
What the quiz screens mean — the timer, badges, reflection, and how your score moves.
Each question counts down from its own target time (set per question, not a fixed 90 seconds). The ring and digits shift from the brand color to amber at 50% time left, then red at 25%, with an urgent pulse at 30 seconds or less and a faster pulse at 10 seconds or less. In Relaxed Mode the timer is informational only: it keeps counting into "Overtime" after zero, and your answer still counts normally. In Strict Mode the countdown keeps running while you are on the question (including if you switch tabs) and auto-submits (or skips, if blank) the moment it hits zero.
You pick this in setup, Step 6 "Time Discipline". Strict Mode auto-submits your answer when the timer hits zero — real exam pressure. Relaxed Mode treats the timer as a suggestion; you can keep working into overtime. Your actual time is still measured either way and feeds the adaptive engine, so time discipline matters even in Relaxed Mode.
A "Revision" chip means you have seen this exact question in an earlier session. It was selected again through the weak/medium/strong or strict-subtopic path (including due spaced repeats from a miss or skip). This is different from Review miss / skip / win chips, which label exhaustion fallback repeats by your last outcome. Getting it right now counts the same as a first-time correct answer: +6 to the subtopic score.
Chips next to each question tell you where it came from. New: never seen before. Discovery: from an untouched subtopic the engine is probing. Review miss / Review skip / Review win: a repeat from the exhaustion fallback path, labeled by your last outcome. Revision: seen in a prior session and selected again on a quota or strict-subtopic path. A question only appears twice when the engine has a reason — usually your own weak spot.
The progress bar and the Correct / Wrong / Skipped tiles update as you answer. They reflect this session only. Your overall mastery is computed per subtopic and shown on the dashboard.
After an incorrect answer you are asked what slipped ("Silly Calculation Mistake", "Forgot Concept/Formula", and so on). Tags are required before you can complete the session — you can still leave without finishing. They teach the engine which kind of error you make, so it can pick better questions. Pick honestly; there is no wrong tag.
Every subtopic starts at a base score of 50 (out of 100). A correct answer adds 6 points, a wrong answer subtracts 8, and a skip subtracts 2. The score is clamped between 0 and 100. Your effective score then decays by 0.5 points per day since you last practiced that subtopic, capped at a 25-point penalty. The engine prioritizes the subtopics where your effective score is weakest.
Skipping counts as −2 to the subtopic score and schedules the question to return for review — it is better than guessing blind. Press S or the Skip button.
Bookmark saves a question so you can find it later. Report flags a question for review if you believe the answer, wording, or explanation is wrong — an admin will check it and you will see the resolution note.