Is There Really a Divide in Ability Between an ATAR of 89.95 and 90.00?

Geoffrey Chen

When Numbers Start Making Decisions · Season One, “People on Either Side of a Threshold” · Article 3

1. How much can 0.05 be worth?

Imagine two applicants who both intend to enter the University of Sydney’s Bachelor of Pharmacy (Honours) and Master of Pharmacy Practice. Both also meet the scholarship’s other conditions, including the citizenship or permanent residency requirement.

The first applicant has an ATAR of 89.95. The second has an ATAR of 90.00.

The University of Sydney currently lists an “ATAR of 90 or above” among the eligibility requirements for the Umberto Cincotta Scholarship in Pharmacy. Eligible applicants are then assessed further on academic merit and a personal statement demonstrating their commitment to studying and practising pharmacy. Reaching 90.00 does not automatically win the scholarship, but an applicant must first cross 90.00 to enter that later comparison. A full-time recipient receives $6,000 per year for up to four years. University of Sydney: “Umberto Cincotta Scholarship in Pharmacy”

In this example, 89.95 and 90.00 are separated by only one reported ATAR band. The second applicant can proceed to assessment; the first is, in principle, ineligible. Their course combinations, commitment to pharmacy, caring responsibilities, school resources and condition on the day of an examination may all differ. Yet the eligibility system first asks only one question: did the number reach 90.00?

This does not mean that every applicant with 90.00 receives the scholarship. It means that 90.00 operates as an entry threshold. Once an applicant crosses it, the personal statement and other academic evidence can be read. If the applicant does not cross it, those materials cannot alter the initial eligibility decision.

The question is therefore no longer only whether the ATAR has been calculated accurately. It is this:

Why can a number designed to rank students’ relative positions suddenly become a line of eligibility at one particular point?

If there is no clear discontinuity in ability between 89.95 and 90.00, why does an institution still need to put two closely placed people on opposite sides of a threshold?

2. The ATAR is a rank, not an examination mark

To understand the threshold, we must first stop reading the ATAR as a percentage score.

The Universities Admissions Centre states explicitly that the ATAR is a rank, not a mark or percentage. Expressed as a number between 0.00 and 99.95, it indicates a student’s position relative to their age group. In New South Wales, the comparison is with the relevant age cohort, not merely with students who completed the HSC or received an ATAR. An ATAR of 90.00 means, approximately, that the student’s overall academic achievement places them ahead of 90 per cent of their age group—in roughly the top 10 per cent. UAC: “What is the ATAR?”

This also explains why the average among students who actually receive an ATAR is usually above 50.00. Not everyone in the age cohort completes Year 12 and obtains an ATAR. When the students who remain in the ATAR-receiving group are, on average, more academically successful than those who left earlier, their average rank moves upwards.

In New South Wales, the ATAR is based on an aggregate of scaled marks from ten HSC units: the best two units of English and the best eight units from the applicant’s remaining courses. Students can choose from many different combinations of subjects, and raw results from unlike courses cannot simply be added as if they measured the same thing. UAC therefore scales course results. The aim is to estimate how students’ results in different courses can be placed on a common scale when the courses are taken by groups with different patterns of overall achievement. Those scaled marks are then used to form an aggregate and a rank. UAC: “How your ATAR is calculated”

At least three transformations occur:

  1. school assessment and external examination results are combined and reported as course results;
  2. results from different courses are scaled to create an academic aggregate that can be compared; and
  3. the aggregate is placed within the relevant cohort and converted into a relative rank.

An ATAR of 90.00 therefore does not mean “90 per cent of the questions were answered correctly”. Nor does 89.95 mean “0.05 per cent less content was mastered”. They are adjacent reporting bands within a ranking system. Each ATAR corresponds to a range of scaled aggregates, and the number of students assigned to each band is also affected by ties in the aggregate.

UAC’s October 2025 technical report confirms that ATARs are reported in increments of 0.05. It also explains that candidates are allocated progressively from the highest aggregate to ATAR bands designed to meet cumulative target frequencies. If students are tied at the minimum aggregate for a band, all students on that tied aggregate are allocated to the ATAR below so the target frequency is not exceeded. UAC: “Calculating the Australian Tertiary Admission Rank in New South Wales: A Technical Report”

The boundary between 89.95 and 90.00 is therefore not a natural scale that cuts knowledge or ability with scientific precision. It is a reporting boundary produced after complex results have been scaled, ordered and grouped.

3. Adjacent ranks do not prove equal ability—but neither do they prove a discontinuity

Saying that there is no discontinuity in ability between 89.95 and 90.00 does not mean that the two applicants are identical in every respect.

Their scaled aggregates and relative positions may genuinely differ. The value of the ATAR lies precisely in its ability to establish an evidence-based order among a very large number of students. When universities have limited places, they cannot pretend that all applicants have exactly the same level of academic preparation.

But the ability to order candidates and the discovery of an essential divide are still two different things.

The ATAR summarises overall senior secondary academic achievement. It does not directly measure a student’s interest in pharmacy, patience in a laboratory, communication skills, long-term discipline, experience of illness, caring responsibilities or the support that will become available after they enter university. It cannot prove that the person in one reporting band will inevitably perform better in a particular degree than the person in the adjacent band.

UAC’s own work concludes that the ATAR is an effective predictor of first-year university performance. In aggregate, groups with higher ATARs are more likely to achieve higher first-year grade point averages. But UAC also emphasises that different courses require different abilities and skills, and that some relevant attributes are not measured by the ATAR alone. The ATAR is one measure of potential, not a universal evaluation of a person. UAC: “The Effectiveness and Limitations of the ATAR”

Statistical prediction normally describes a relationship across groups. Even if students in the 90s have a higher average university performance than students in the 80s, it does not follow that every student with 90.00 will outperform every student with 89.95. It follows even less that 90.00 is exactly where “having the ability” begins and 89.95 is exactly where “not having the ability” ends.

The most accurate conclusion is therefore this: the ATAR can provide evidence about academic preparation, but a small ordering difference between adjacent bands is not sufficient to prove a discrete change in a person’s ability.

4. Universities often need a stopping point, not a boundary in ability

If 90.00 is not a natural boundary in ability, does that make the threshold unreasonable?

Not necessarily.

When places, scholarships and assessment time are limited, an institution must decide where to stop. Suppose a course has one hundred places and five hundred qualified applicants. Even if every applicant is capable of completing the course, the university cannot issue an offer to everyone. Ranking in this context may be less about separating those who can succeed from those who cannot, and more about deciding who receives a scarce opportunity first.

This reveals why admissions numbers are so often misread. A course threshold may look like a pure academic standard, while actually combining at least three different considerations:

  • the minimum preparation needed to complete the course;
  • the relative competitiveness of that year’s applicants; and
  • the number of places the university can provide.

If the second or third factor changes, the position of the admissions boundary may change even when the difficulty of the course remains the same.

UAC’s published “lowest selection rank” is generally the selection rank of the last Year 12 applicant to receive an offer for a course in a particular round. It is a result of competition, not necessarily a pre-established minimum ability standard. It is also not necessarily the lowest raw ATAR among those who received offers. If application numbers, available places or the applicant distribution change in the following year, the lowest selection rank may rise or fall. UAC: “December Round 2 lowest selection ranks”

A fixed scholarship requirement of 90.00 operates differently. It is not the rank of the last recipient discovered after all applications have been compared. It is a condition established in advance. It can reduce the number of applications requiring detailed assessment and can communicate the level of academic achievement expected by the scholarship. But precisely because it is a line selected beforehand, the institution assumes an explanatory burden: why 90.00 rather than 89.00 or 85.00? Why use a fixed ATAR threshold instead of comparing all otherwise eligible applicants through a combined assessment?

“A line must be drawn somewhere” explains why a threshold may exist. It does not, by itself, explain why the line has been drawn here.

5. The decisive number in admissions is often no longer the ATAR

When students see that a course has a “minimum of 90”, they may assume that everyone must have a raw ATAR of at least 90.00. Actual admissions systems are often more complicated.

Most current Year 12 applicants compete using a selection rank. The selection rank is formed by adding applicable adjustments to the ATAR. Those adjustments may recognise strong performance in relevant HSC subjects, a student’s home or school location, or disadvantage accepted under the Educational Access Scheme. The adjustments do not change the ATAR. They change the applicant’s position for a particular course preference, and the same person can therefore have different selection ranks for different courses. UAC: “University selection rank adjustments”

An applicant with an ATAR of 89.95 may, after subject or location adjustments, have a selection rank above 90.00 for one course. Another applicant with an ATAR of 90.00 and no adjustments may retain a selection rank of 90.00. For that course, the first applicant may be placed ahead of the second.

This does not mean that every rule written as “ATAR 90” accepts an adjusted selection rank. Scholarships, guaranteed entry schemes, special programs and ordinary course competition can each use different definitions. Some rules expressly require a raw ATAR; some use an adjusted selection rank; others require a minimum ATAR before any adjustments can be applied.

Every number with decision-making power should therefore answer at least four questions:

  • Does the page mean ATAR or selection rank?
  • Do adjustments apply, and are they capped?
  • Is 90.00 a guaranteed outcome, a minimum eligibility condition or the result of an earlier admissions round?
  • Are there prerequisites, an interview, a portfolio or a personal statement in addition to the rank?

If these distinctions are not explained clearly, the student appears to face a precise threshold but does not know which number the institution will actually use.

Adjustment mechanisms are intended to supplement relevant information that the ATAR omits. Yet they create a new obligation of explanation. A public raw rank combined with an opaque set of adjustments can generate a decision that may better reflect a policy goal but is harder for the applicant to understand. Fairness requires more than considering additional factors. It also requires applicants to know which kinds of information affected their result.

6. A calculation error can be corrected—but how should a hard threshold be reviewed?

When a number determines access to an opportunity, two different kinds of dispute must be separated.

The first is an error in data or calculation. An examination result may be amended, a course unit may be omitted, or application information may not have been matched correctly. UAC states that if the NSW Education Standards Authority provides amended HSC results, the ATAR is automatically recalculated and the student is notified. UAC also provides channels for questions about calculation around the time ATARs are released. UAC: “How your ATAR is calculated”

The purpose of this form of correction is to make the system produce the right number under its existing rules.

The second dispute is harder. The number has been calculated correctly, but has the rule placed too much weight on it? A student with 89.95 cannot demand that UAC raise the rank merely because it is close to 90.00. The ATAR may be entirely accurate. The real object of review is the institution’s decision to give two adjacent reporting bands sharply different consequences.

For competitive places, review cannot promise every borderline applicant a place. As long as resources are limited, someone will remain outside the final boundary. Institutions can, however, reduce the risk of disguising an allocation order as a judgement of ability:

  • state whether the threshold is a minimum preparation standard, an administrative screening line or a stopping point created by competition;
  • distinguish clearly among raw ATAR, selection rank and a historical lowest selection rank;
  • for programs that rely heavily on a single score and have significant consequences, consider supplementary review for applicants very close to the threshold;
  • where the relevant goal involves particular capabilities, include subjects, portfolios, interviews or experience that measure those capabilities more directly, rather than asking an overall rank to substitute for every judgement;
  • explain the basic logic of adjustments and allow applicants to correct information about location, educational history and special circumstances;
  • compare the threshold periodically with students’ later university performance to test whether the line still serves its original purpose; and
  • preserve pathways through transfer, enabling programs or later reapplication so one boundary decision does not become a permanent identity.

UAC calculates the rank and processes applications, but each university ultimately decides which applicants receive offers. A query about the ATAR calculation, a correction to application data and a review of a university decision belong to different institutions and procedures. Unless responsibility is made clear, applicants cannot know who should be asked which question.

Conclusion: 0.05 can decide eligibility, but it cannot prove a sudden change in ability

The ATAR addresses a real problem. Students complete thousands of different combinations of courses, while universities must compare applicants and allocate limited opportunities within a limited period. Without a common ranking tool, decisions could become less transparent and more dependent on connections, impressions or untested subjective judgements.

A reasonable criticism is therefore not that all rankings should be abolished, or that 89.95 and 90.00 contain absolutely no statistical difference.

What must be resisted is the leap from order to essence: because one person is above a threshold, the institution appears to have proved that the person “has the ability”; because another is one band below, it appears to have proved that the person is “not good enough”.

Within a particular scholarship, 90.00 can operate as an entry condition that manages applications and communicates an academic expectation. It remains an institutional choice, not a cliff growing naturally out of human ability. It can determine who enters the next stage without discovering two fundamentally different kinds of person.

The most accurate account of the ATAR is therefore:

It can help institutions order access to opportunities, but it should not be used to define the whole of a person’s potential.

The greater a number’s apparent precision, the easier it is for a threshold to look objective and beyond argument. Yet the most important fact about the interval between 89.95 and 90.00 is not the decimal precision. It is that a decision-maker has chosen this point as the place where consequences change.

Once a number acquires power, the obligation to explain must grow with it: what does the number measure, what does it omit, why is the threshold here, is there supplementary judgement near the line, can errors be corrected, and does one unsuccessful decision leave other routes open?

Only under those conditions can ranking remain an acceptable tool of allocation rather than a verdict of destiny wearing the costume of a decimal.

Primary sources

Discover more from Geoffrey Chen

Subscribe to get the latest posts sent to your email.