We publish our ranking rules so teachers can focus on teaching and parents can trust the results.
All metrics only count the last 30 days，How long a teacher has been on the platform does not affect ranking.
Each dimension is calculated as students meeting the condition divided by all students of the teacher, so inactive names do not add advantage.
Score improvement and high-scoring student ratio together account for 60% of the weight.
Updated automatically every minute — the latest teaching results show up on the leaderboard right away.
The top 20 are sorted by overall score; from rank 21 onward, teachers are shown randomly to give every listed teacher fair exposure.
| Group | Dimension | Weight | Info |
|---|---|---|---|
| Effectiveness |
Share of students with score improvement
%
|
35% | Students with at least one piece whose highest score exceeds its first score in the last 30 days ÷ all students of the teacher |
| Effectiveness |
Share of students scoring 90+
%
|
25% | Students with at least one analysis report scoring 90+ in the last 30 days ÷ all students of the teacher |
| Output |
Share of students with 5+ reports
%
|
25% | Students with at least 5 sheet-music analysis reports in the last 30 days ÷ all students of the teacher |
| Activity |
Share of students with uninterrupted recent activity
%
|
15% | Within the last 4 weeks, students whose activity is uninterrupted when counted backward from the most recent week; 1 to 4 consecutive weeks all count ÷ all students of the teacher |
| Total | 0% | ||
ranking_score = Σ ( weight_i × percentile_i ) × 100
Each dimension is 0~100% student ratio, mapped to a percentile (0-1) across listed teachers and weighted by the table above to produce a final 0-100 score.
The denominator is all students bound to the teacher, not only active students, so having more low-participation students does not create an advantage.
Rules may be adjusted based on platform operations; this page is updated whenever they change.