Question 1

How accurate is the Uni Predictor?

Accepted Answer

Each estimate carries an 80% confidence band that reflects cycle-to-cycle variation and factors public data does not capture. We start from published offer rates, then shift in log-odds space using factors admissions data shows are predictive. The model recalibrates whenever new End-of-Cycle data or FOI returns arrive. We never claim to predict your actual outcome, only to bound it with honest uncertainty.

Question 2

What data is the Uni Predictor built on?

Accepted Answer

UCAS End-of-Cycle reports, FOI returns from Cambridge, Imperial and Oxford, Cambridge Undergraduate Admissions Statistics 2022 to 2024, Oxford Mathematical Institute and CS Feedback PDFs, Imperial transparency workbook, LSE Planning Division dashboards, UCL Apps-per-Place data, OfS contextual admissions framework, Sutton Trust personal statement research, and Cambridge college-level subject offer rates.

Question 3

Does TMUA score really matter for Cambridge?

Accepted Answer

Yes. TMUA is treated as a heavyweight factor by Cambridge tutors despite the absence of a published threshold. Our TMUA coefficient draws on direct conversations with current Cambridge maths tutors and recent admitted students. Cambridge actively refuses to publish a TMUA-to-offer threshold, so we calibrate to documented cohort ranges plus contemporaneous tutor signal. As Cambridge Assessment publishes more TMUA outcome data, we recalibrate.

Question 4

What is the average TMUA score for a Cambridge offer?

Accepted Answer

No university publishes a hard TMUA threshold, but successful Cambridge applicants for Maths, Computer Science and Economics typically sit in the 6.5 to 7.5 band on the 2025 scale (around the 80th to 95th percentile). Imperial Computing publishes a clearer cutoff via its TMUA dashboard. Use the predictor to see your specific course chance at your specific score.

Question 5

Can I get into Cambridge with AAB predictions?

Accepted Answer

It is harder but not impossible. AAB sits below the A*A*A baseline, so the grade-effect adjustment is negative. Around 94.7% of UK Cambridge accepts achieved A*AA or higher (2022 to 2024 cycles). However, a strong TMUA or STEP performance can partially compensate, and contextual flags can lift probability by 7 to 15 percentage points. Run your specific profile through the predictor to see the calibrated chance.

Question 6

How does the Cambridge pool affect my chances?

Accepted Answer

Around 700 Cambridge applicants are pooled each year. Around 22% of them receive an offer from a different college. Our model computes combined probability as P(direct offer) plus P(rejected directly, pooled, then pool offer). Pool entry probability peaks around the direct-offer rate, so very strong and very weak applicants pool less often. Cambridge Maths pools at around 25%, Computer Science 22%, Engineering 21%, Economics 18% (2024 cycle averages).

Question 7

What is the difference between Trinity and Robinson for Cambridge Maths?

Accepted Answer

Trinity Maths offer rate sits at around 18.6% versus Robinson Maths at around 33.9% for an equivalent applicant profile (Cambridge Admissions Statistics 2024). Trinity attracts a stronger applicant pool with very high STEP and TMUA averages. The predictor applies real per-subject offer-rate multipliers from college-level data, so switching college in the tool will show you the calibrated delta.

Question 8

Why is Churchill Computer Science offer rate so low?

Accepted Answer

Churchill is a CS-heavy college that draws a disproportionately strong applicant field. The headline offer rate is suppressed because the denominator is large and weighted with very high-quality applicants. The predictor adjusts for this with real per-college, per-subject multipliers rather than naive offer-rate averages.

Question 9

How does the predictor handle STEP for Cambridge Maths?

Accepted Answer

STEP is the largest binding constraint for Cambridge Maths. About 48% of Cambridge Maths offer-holders convert to a place each cycle, with STEP grade misses driving most of the gap. STEP grade 1 carries a +0.26 log-odds shift. STEP S adds +0.79. Without a STEP plan in your profile, the model treats this as a major outstanding factor and widens the confidence band accordingly.

Question 10

Does the predictor account for state versus independent school?

Accepted Answer

Yes. We use the Boliver (2013) odds ratio of 0.59 for state versus independent at equivalent grades, blended with each university's own grade-controlled gap from OfS 2024 data. For Cambridge specifically, grammar-school applicants accepted at 24.4% in 2024, ahead of independent at 21.6%. The predictor reflects these real numbers rather than assumed disadvantage.

Question 11

Is the Uni Predictor really free?

Accepted Answer

Yes. The predictor is fully free, no sign-up, no email gate, no paywall. We build it because honest probability matters and the existing tools were not built on real data. TMUA.co.uk offers paid TMUA prep elsewhere on the site, but the predictor itself stays free and ungated.

Question 12

How is this different from a UCAS Tariff calculator?

Accepted Answer

A UCAS Tariff calculator only adds up your UCAS points. The Uni Predictor models offer probability: it takes your grades, admissions test score, school type, college choice, contextual flag and predicted-vs-achieved status, then produces a calibrated percentage chance of an offer at your chosen course. Tariff points are an input. Offer probability is the output.

Uni Predictor

tmua is your biggest lever.

This scenario ranks above 80% of modelled mathematics profiles.

Lucy Cavendish beats Trinity by 22%.

92% of the 2025 reference cohort scored below 6.5.

+36 points in this model if TMUA rises to 9.0.