The graph gives order, not pacing. Our moat layers: difficulty calibration, time-on-task, retention/spacing, motivation signals, edge validity.
Decide: which layer to instrument and calibrate first, and the minimum-data bar before it influences path decisions.
Suggested default: difficulty (IRT per topic) — it feeds assessment quality immediately and every other layer benefits from it.
Context: 01-education-learning-sciences.md, Q1.2
The graph gives order, not pacing. Our moat layers: difficulty calibration, time-on-task, retention/spacing, motivation signals, edge validity.
Decide: which layer to instrument and calibrate first, and the minimum-data bar before it influences path decisions.
Suggested default: difficulty (IRT per topic) — it feeds assessment quality immediately and every other layer benefits from it.
Context: 01-education-learning-sciences.md, Q1.2