How to read your band gap and know what to fix first

Every skill on your dashboard shows a current estimate next to your target band, plus a status dot: green means you're at or above target, amber means you're close, red means there's real distance to close. The number itself isn't a single test score — it's a rolling estimate built from your recent attempts in that skill, weighted so that more recent, more consistent performance counts for more than a single old result. That's a deliberate design choice, not a limitation: a single Reading passage or one Speaking session is a noisy signal on its own (anyone can have an unusually good or unusually bad ten minutes), while a trend across several attempts is a far more honest picture of where you actually stand.

The instinct most candidates have is to practice whatever they're worst at, and that instinct is usually right — but not always. If your worst-looking skill has barely any data yet (one or two logged attempts, or none), the estimate for it is genuinely unreliable, closer to a placeholder than a measurement. A skill showing a real, well-supported gap based on several consistent attempts is a far more trustworthy signal to act on than a single rough attempt in a different skill that happened to land badly. Before reorganising your whole study plan around whatever looks reddest today, check how many attempts that number is actually built from — a red dot with three attempts behind it deserves more weight than a red dot with one.

This is exactly the judgement call "Biggest opportunity" is designed to make for you automatically. Rather than simply pointing at your lowest raw score, it looks for the skill with a genuine, well-supported gap and then goes one level deeper: it names the specific competency inside that skill actually dragging the average down — a particular Reading question type you consistently lose marks on, a particular Listening section pattern, or a Writing criterion like Coherence and Cohesion or Lexical Resource specifically, rather than "Writing" as an undifferentiated whole. That specificity matters because two candidates can show an identical overall Writing score for completely different reasons — one losing marks on task achievement, the other on grammatical range — and the right next practice session looks nothing alike for the two of them.

It's worth understanding why the estimate can occasionally feel out of step with how a session actually felt. A single strong attempt right after a long run of weaker ones won't instantly flip a red dot to green, because the estimate is deliberately resistant to being swung by one data point in either direction — the same stability that stops a single bad day from tanking your number also stops a single good day from inflating it. If you've just had what felt like a breakthrough session and the dashboard hasn't caught up yet, that's the system working as intended, not a bug: give it two or three more consistent attempts in that skill before expecting the estimate to visibly shift.

Practical rule: once "Biggest opportunity" points you somewhere, commit to working that specific competency for at least a week before switching, then check whether the gap actually narrowed before deciding what's next. Chasing a new "weakest skill" every session, or bouncing between skills because one felt discouraging, means you never spend enough consecutive, focused time on any single weakness to actually move it — and a gap that gets a little attention from five different directions closes slower than one that gets sustained attention from one.