The dashboard is where every question you log turns into something you can act on. But a number is only useful if you know exactly what went into it. This post is the reference for that: first how to log so the analytics are meaningful, then precisely how each widget is calculated - with a full breakdown of the one number students always ask about, your syllabus pace.
Everything below describes the actual math the app runs. Nothing here is an estimate or an analogy.
First: logging is the input, the dashboard is only the output
The dashboard can only ever be as good as what you record on the log screen. After a question ends (either you tapped the timer early because you solved it, or the timer hit zero), the log screen captures a small set of fields. Each one feeds a specific part of the dashboard:
- Solved - yes / no. This is the single most important field. It drives your solve count, success rate, the accuracy trend, and the weak-chapter ranking. A question with no solved/no value recorded is skipped entirely by accuracy math.
- Difficulty - easy / medium / hard. Feeds the Question Quality breakdown. Without it, that chart has nothing to show.
- Mistake type (only asked when you didn't solve) - silly, conceptual, mis-identification, formula, calculation. Feeds the Mistake Breakdown pie and the "mistakes" KPI.
- Resource, notes, the wrong option - not charted, but stored on the question and available to the Advisor for exact lookups.
- Time on the question - recorded automatically. Feeds average time, the speed trend, and the session-length stats.
The practical rule: always set solved, always set difficulty, and always set a mistake type when you miss one. Those three take two taps and unlock most of the dashboard. Skipping them doesn't break anything - the widgets just quietly have less to say.
How the dashboard is organised
At the top is a filter row with two dropdowns: subject, then chapter. Everything below reacts to them. Pick a subject and every widget re-scopes to that subject's questions; pick a chapter and it narrows further. Leave both on "All" to see the whole picture.
Several widgets also have their own Overview / Today toggle. Overview uses everything in the current filter; Today restricts that widget to questions logged on the current date. The toggle is per-widget - flipping the accuracy chart to Today doesn't touch the others.
The numbers up top (KPI cards)
These are straight aggregates over whatever the filter currently resolves to:
- Questions - the count of logged questions in view.
- Solved - how many of those are marked solved. The little bar underneath fills to solved ÷ total.
- Success rate -
round(solved ÷ total × 100). It's colour-coded: green at 70% or above, amber from 40–69%, red below 40%. - Avg time / question - the mean of every question's recorded time, shown as minutes:seconds.
- Mistakes - how many questions have a mistake type recorded. Green when that's zero; amber while it stays under a fifth of your questions (specifically
max(2, 20% of total)); red above that.
Session-length cards
A "session" here means one chapter's practice run - the questions grouped under a single chapter, not a wall-clock block. From that grouping:
- Avg questions / session - total questions ÷ number of chapters that have any logged questions.
- Avg session length - the average, across chapters, of the summed per-question time in that chapter.
- Longest session - the largest of those summed per-chapter times.
Because these sum the per-question timer values, they measure time actually spent attempting questions, not idle time with the app open.
The charts over time
Activity heatmap - a GitHub-style grid of the last fourteen weeks (it starts on the Sunday on or before 90 days ago, so weeks line up as clean columns). Each cell is one day. A day with no questions is blank; otherwise its shade is ceil(count ÷ busiest-day-count × 4), capped at four levels. So the darkest cells are always relative to your busiest day in the window, not a fixed target. Hover any cell for the date and exact count.
Activity line - questions logged per day across the last 30 days. This is raw volume, nothing weighted.
Accuracy trend - this is a running accuracy, not a per-day one. The app walks your questions in the order they were logged, and after each answered question plots your cumulative success rate so far: round(running solved ÷ questions answered so far × 100). Only questions with a solved yes/no count; anything left blank is ignored. Early on the line swings a lot (a single miss moves it several points); as volume grows it settles, which is exactly what a cumulative average should do. The Today toggle restricts it to just today's questions.
Speed trend - a rolling average of your last 10 questions' times, plotted once per question. Each point is the mean time-per-question over that trailing window, in seconds. A line trending down means you're getting faster over recent questions; a spike means a run of slow ones. It's paired visually with accuracy so you can see the trade-off - fast and accurate is the goal, fast but sinking accuracy is a warning.
The breakdowns
Mistake Breakdown - of the questions where you recorded a mistake type, this splits them by type: silly, conceptual, mis-ID, formula, calculation. It only counts questions that actually have a mistake type, so it's answering "when I get things wrong, why" - not "how often am I wrong."
Question Quality - of the questions where you recorded a difficulty, this splits them easy / medium / hard. Note this is about the difficulty mix you're practising, independent of whether you solved them. A wall of "easy" is a signal you may be practising in your comfort zone.
Weak Chapters - the leaderboard of chapters you're struggling with most. A chapter qualifies once it has at least 3 answered questions (so a single unlucky miss can't top the list), and chapters are ranked by lowest accuracy - solved ÷ answered - showing the six weakest. Each row is colour-toned the same way as the success KPI (green ≥70%, amber 40–69%, red below). This is your "what to revise next" shortlist.
Syllabus pacing - how your pace is actually defined
This is the widget worth understanding in full, because "am I on track?" is the whole reason to track pacing, and the honest answer depends entirely on what "pace" is measuring.
Pacing only appears once you've set a study goal. You do that from the account menu (the card at the bottom of the sidebar) → Set study goal - not through the Advisor. You give it a short description and a duration (e.g. "3 months" or "90 days"), which the app turns into a deadline date.
Once that's set, here's every input the gauge uses.
Coverage - the gauge percentage. Every chapter you've entered is sorted into one of three buckets by how many questions you've logged against it:
- Untouched - zero logged questions.
- Light - 1 to 4 questions.
- Solid - 5 or more questions.
The big number in the ring is your coverage: round((total chapters − untouched chapters) ÷ total chapters × 100). In plain terms, it's the share of your chapters that have at least one logged question.
Required pace. The app counts the days left until your deadline (whole days, floored at zero), then:
required pace = untouched chapters ÷ days remaining
That's how many fresh chapters you'd need to start each day to have touched every remaining chapter by your deadline. If the deadline has passed and chapters are still untouched, it shows "unreachable." If nothing is untouched, required pace is zero.
Actual (current) pace - and this is the part to read carefully. Your current pace is:
actual pace = (chapters whose first-ever logged question falls within the last 14 days) ÷ 14
Read that definition closely, because it decides what "keeping up" means:
- It counts a chapter once, on the day you first opened it. Grinding a chapter you started three weeks ago does not add to your current pace - that chapter was already "covered."
- It's therefore a measure of how fast you're opening new ground, not how many questions you're doing or how well you're doing them. A 200-question day spread across chapters you'd already started registers as zero new pace.
- The window is a fixed trailing 14 days, so it reflects your recent breadth, not your all-time average.
On track or behind. The gauge turns green when you're on track and red when you're behind, where on-track means: required pace is zero (nothing left to cover), or your actual pace is at least the required pace.
The limitation this creates - and why you should read pacing alongside other widgets
Because pace is built on coverage (did a chapter get a first question) rather than mastery (are you actually good at it), there's a blind spot worth naming plainly:
Once every chapter has at least one logged question, coverage hits 100%, untouched drops to zero, required pace becomes zero - and the gauge will say "on track" forever, even if you've done exactly one shaky question in each.
That's not a bug; it's what "have I at least started everything by my deadline?" honestly answers. But it means pacing alone can make you feel finished when you've only skimmed the surface. So pair it:
- Pacing tells you whether you've reached every chapter in time (breadth).
- Weak Chapters and Success rate tell you whether you're actually good at them (depth).
- Question Quality tells you whether the breadth you built was on real difficulty or easy reps.
A student who's "on pace" with a weak-chapters list full of red isn't done - they're covered but not competent, and those two widgets side by side are how you catch it.
(One guardrail the app adds: if you've entered fewer than 10 chapters total, the pacing figure is flagged as a rough signal, since it can only reason about chapters you've actually added, not the full official syllabus.)
Putting it together
A quick weekly read that uses the dashboard the way it's built to be used:
- Coverage on track? Glance at the pacing ring - are you reaching new chapters fast enough for your deadline?
- Depth holding up? Scan Weak Chapters. Anything red is next week's revision, regardless of what pacing says.
- Honest difficulty? Check Question Quality. If it's mostly easy, your success rate is flattering you.
- Speed vs accuracy? Look at the speed and accuracy trends together - getting faster is only good if accuracy holds.
Log the three key fields on every question, and all four of those reads take about thirty seconds. That's the whole point of turning each question into a data point: not to admire the charts, but to know precisely what to do next.