Adaptive Practice
Retention comes from choosing the next problem for the right reason
Random worksheets do not build durable skill. MathRoutine uses recent attempts, topic weakness, skill tags, pacing, and higher-order reasoning signals to focus practice where it can actually change the student.
Product Evidence
Weakness Tracking
Topic weakness is broken into smaller skill causes
Reports separate topic performance from skills such as equation setup, comparison, fraction reasoning, unit conversion, and two-step reasoning.
Reasoning Signals
Higher-order blockers are tracked directly
Upper-grade practice can surface hidden quantity, reverse reasoning, distractor information, comparison logic, and multi-constraint modeling.
Focused Mode
Recommended practice can start from the weakest pattern
The weakness engine can send a student back into practice with the grade, topic, difficulty, and reasoning marker that need attention.
Tutoring Flow Example
The important part happens after the student tries
MathRoutine is designed to separate the wrong answer from the reason behind the wrong answer, then make the next practice narrower and more useful.
Problem
A school sold 80 tickets. Adult tickets cost $12 and student tickets cost $7. The total collected was $760. How many adult tickets were sold?
Student Response
The student tries one equation using only the money total and guesses the number of adult tickets.
Diagnosis
The issue is not arithmetic first. The student needs a two-relationship model: total tickets and total money.
Intervention
Lower the scaffold for one attempt: define a and s, write the count equation, then write the revenue equation.
Next Practice
Follow with another systems problem that changes the story surface but keeps the two-constraint structure.