What match score and admission likelihood mean

The Match for me tab of the university catalogue matches programmes to your IB profile. Every result there carries two judgements: a match score, the number in the ring, and an admission likelihood, the coloured label. They answer different questions, and they are worked out in different ways.

What the score is

The match score says how well a programme fits you. The higher the number, the better the fit. It weighs your IB results against the programme's requirements, the programme's field against your interests, the country and the teaching language, and the cost.

The likelihood answers a narrower question: are the programme's entry requirements within reach of your predicted results? A programme can fit you well and still be a reach. It can also be well within reach and fit you poorly. So a high score next to a cautious likelihood isn't a contradiction.

The four likelihoods

Each result shows one of these: Strong, Target, Reach, Unlikely.

The likelihood comes from a fixed rule, not from an AI model. Accepted compares your predicted IB total with the minimum score the catalogue holds for the programme:

Strong when your total is a few points or more above the minimum;
Target when you meet the minimum, or are less than a few points above it;
Reach when you are up to a few points below it;
Unlikely when you are further below it.

Three things change that reading:

Subjects come first. If the programme requires a subject you don't take at the level it asks for, the likelihood is Unlikely, whatever your total. If you take it at the right level but your predicted grade is below the one asked, the likelihood is Reach at best.
An uncited minimum is never the lowest band. Universities often publish a range, and a single figure can't show that. When the catalogue has no source recorded for a programme's minimum, the likelihood stops at Reach.
No minimum, no comparison. When the catalogue holds no minimum score for a programme, or your profile has no predicted total, there is nothing to compare. The likelihood then shows Target unless a subject lowers it. When the minimum is missing, the result's list of requirements says so, and that the band is only indicative.

What it is based on

Which programmes. The matcher needs your IB profile with at least one subject. From Accepted's catalogue it takes the programmes closest to your subjects and interests, within the fields and countries you chose. Those come from the catalogue's filters, or from your profile when no filter is set. If your profile sets a yearly tuition limit, programmes known to cost more are left out. Programmes you already applied to, and those of universities you saved as favourites, are always added.

The score. An AI model judges the score. It is given your profile and what the catalogue holds about each programme, and asked to weigh academic fit, how the field fits your interests, the country and language, and the cost. It also sees the applications you already have, and leans towards new programmes in the same country and field. The model writes the short explanation on each result too.

The fixed limits. After the model has scored, Accepted applies fixed caps:

if the programme requires a subject you don't take at the level it asks for, the score can't go above a low ceiling;
if your predicted grade in a required subject is below the one asked, the score can't go above a higher ceiling;
the language caps apply only to a programme that isn't taught in English and for which the catalogue holds a required level in its teaching language. If your profile lists your languages and that language isn't among them, the low ceiling applies. If you listed it with a level below the one required, the higher ceiling applies. If your profile lists no languages, or lists that language without a level, there is no language cap.

The lists. The requirements met and missing under each result come from the catalogue and your profile, not from the AI model. A check also removes sentences of the explanation that contradict your standing against the entry minimum, a required subject or a language level.

When the AI model doesn't answer, Accepted scores the programmes with a fixed formula instead: how close each programme is to your profile, and your total against its minimum. The same caps and the same likelihood rule apply.

Saved results. When you come back to an earlier run, Accepted reads the programme facts from the catalogue again, and works out the likelihood and the caps again against your current profile. The AI model's own score and its explanation stay as they were in that run. The saved explanation is checked again against your current standing, and sentences that contradict it are removed. If a programme has since been retired or removed from the catalogue, its result says so and shows no likelihood. If your profile has changed since, the page says so, and running the matcher again gets you fresh explanations.

What it is not

A likelihood is not a university's decision, and a high score is not a promise of a place. Both are read from the data Accepted holds and the profile you entered. Where that data is incomplete, they can only be as good as it is.

The matcher doesn't read your personal statement or your portfolio, and it can't know how an interview will go. When a programme requires a portfolio or an interview, the list of what's missing says so.

Use the results to decide which programmes to look at closely, then check the requirements on the university's own site and talk them through with your counsellor.