Methodology & data sources

Priority points

Applies regional and applicant-priority points, then tapers linearly after 22.5/30. Coefficients live in CONFIG and must be checked against each year's regulations.

Cutoff prediction

Sample data uses recent-year scores and a simple linear trend. Real data will be replaced by a separate ingest pipeline.

Probability

Probability uses a logistic curve from the gap between effective score and predicted cutoff, clamped between 2% and 98%.

Wishlist

P(>=1 admission) = 1 - product(1 - p_i). This independent assumption is for reference and does not reflect the official virtual-filtering process.

Estimates are for reference only and are not official Ministry of Education and Training results. Method →