Adaptive paths: smarter changes, told to your face, and proven
Your path now adapts from every signal - scores, retries, your own too-easy/too-hard feedback, and pacing - with guided review steps when you struggle, a visible history of every change, and a built-in experiment proving adaptation works.
The adaptation layer grew from one rule into a real decision system - and it explains itself.
- Tell us how a chapter felt: a new "Too easy / Just right / Too hard" control feeds your opinion straight into how the rest of your path is shaped - it outranks every inferred signal.
- When checks take repeated tries or a skill assessment shows a gap, the path now inserts a guided step-by-step review of that skill instead of just easing the difficulty.
- Struggling and slow? The next chapter switches to a walkthrough format. Acing checks in record time? The path steps up and says why.
- Every change is explained with its reason - including alongside milestones, which previously hid them - and a permanent "How your path has adapted" history lives on your path view, with changed steps marked "Adjusted" and inserted ones "Added for you".
- Regenerated easier chapters now confirm what changed and keep a full lineage to the original version.
- A small share of learners keep a frozen path as a control group, powering an on-platform readout of whether adaptation genuinely improves pass rates and gap closure - evidence, not just adaptation.