Clinical Problem Solving & Professional Dilemmas
The score band distributions used in this calculator are taken directly from the GP ST1 2026 Round 1 scoring tables published by NHS England on the GP MSRA page. Each table shows what percentage of the GP applicant cohort fell into each standardised score range for the Clinical Problem Solving (CPS) and Professional Dilemmas (PDT) papers. NHS England also confirms that scores are normalised around a mean of 250 with a standard deviation of 40 per paper.
The GP regional comparison data comes from a Freedom of Information request — applicants ranking for GP August 2025 — published on WhatDoTheyKnow. The FOI response contains the actual MSRA scores of every offer made and accepted for each trust in the August 2025 GP ST1 intake, allowing direct comparison of your score against the real distribution of successful applicants by region.
The Core Psychiatry regional comparison data comes from a separate FOI request — MSRA scores by deanery and trust — also published on WhatDoTheyKnow. This response contains MSRA scores of offered and accepted applicants broken down by deanery and trust for the 2025 Core Psychiatry training intake. Unlike the GP data, no score band distribution is publicly available for Psychiatry, so individual percentile estimates and the combined percentile chart are not shown in Psychiatry mode.
Each score band in the published tables represents a histogram bucket — a range of scores with a known proportion of the cohort sitting within it. To estimate your percentile within a given bucket, the calculator assumes your score falls uniformly across that bucket's range.
Concretely: if you score 280, which falls in the 271–290 range (24% of candidates), the calculator first sums all percentages below that bucket, then adds a proportional fraction of the bucket itself based on where 280 sits between 271 and 290. This gives a smooth percentile estimate consistent with the published distribution without overfitting to the coarse histogram.
To estimate means and standard deviations for each paper, the calculator uses the histogram midpoints and their associated probabilities to compute weighted means and variances directly from the published data. This yields:
These are slightly different from the nominal 250/40 values because the actual published distributions are not perfectly symmetric around 250 — the PDT distribution in particular is left-skewed, with more candidates concentrated in the 211–250 range.
GP training ranks candidates on their total MSRA 2026 score (CPS + PDT combined). Simply averaging your two individual percentiles would be misleading because the percentile scale is non-linear — differences between candidates compress toward the extremes.
Instead, the calculator models the two papers as a bivariate normal distribution with a correlation of r = 0.65. This reflects the reasonable assumption that overall candidate ability links performance across both papers — stronger candidates tend to score higher on both. The combined distribution's standard deviation is therefore:
Your combined score is then converted to a z-score relative to the combined mean, and mapped to a percentile using the standard normal CDF. Assuming independence (r = 0) would artificially compress the combined SD and push extreme scores further into the tail than they deserve — hence the correlation correction.
This calculator provides an estimate only. The true inter-paper correlation is not publicly published — r = 0.65 is a reasonable assumption based on typical cognitive assessment correlations, but the true value may differ. Percentile estimates for scores near the boundaries of the published ranges (below 170 or above 310) are extrapolated and should be treated with caution. The distributions reflect the 2026 Round 1 GP applicant cohort only and may vary in future rounds.