The standard advice on a hard dMAT eligibility question is to ask rather than guess, and it is good advice. What it does not prepare you for is what happens on the other side of asking. You put the question to a university's admissions office, to a consultant, to a seniors' group on WhatsApp and to the official contact route, and the replies do not agree. You now feel less certain than you did before asking, and the instinct is to count the replies or to trust whichever sounded most confident. Neither is a method, and neither is necessary, because in most cases the replies were never answers to the same question.
Three bodies, three different questions
The dMAT itself is run by g.a.s.t., which publishes who the test is for, how registration and test dates work, and when certificates are issued. The verification procedure is run by APS India, which publishes which applicants it requires a dMAT certificate from and what a complete file must contain. The programme you want to study is run by a German university, which sets its own admission requirements, intakes and deadlines.
Those are three separate responsibilities, and each source is authoritative over its own and nothing else. A university admissions officer describing the dMAT rule is repeating someone else's rule, accurately or not. APS India does not decide whether a department will admit you. And a consultant who describes all three is, at best, relaying all three. None of that makes them dishonest; it makes them differently placed, which is more useful to know. How the dMAT and APS processes fit together is worth reading once for this reason alone: you cannot weigh an answer until you know which part of the process the answerer owns.
So before you treat a reply as an answer, write down which of the three it came from, and whether the question you asked was actually theirs to settle.
Most conflicts are two answers to two questions
Line the replies up properly and most apparent contradictions dissolve. Four ordinary reasons produce them.
A different question. "Do I need APS?" and "do I need a dMAT certificate as part of APS?" are different questions with different answers, and in a fast conversation they compress into the same sentence. So do "is my degree in an affected field?" and "is my programme in an affected field?" — the requirement, on the official information currently available, attaches to the field of the degree you already hold, not to the Master's you are applying into.
A different applicant. An answer built for a friend or a forum poster carries their degree title, their intake and their registration dates inside it. It may be perfectly correct — about them.
A different date. The requirement was introduced with a start intake and a small set of transitional exemptions attached to specific 2026 dates. An answer given before a published change is not wrong so much as stale, and an answer with no date attached cannot be checked for staleness at all.
A different level of detail. The published field list is a general statement; a ruling on your particular transcript is not, and the first cannot produce the second. This is exactly the gap that interdisciplinary and mixed degrees fall into, where a title sits across category lines and the list simply does not reach far enough to decide it.
The practical test takes ten minutes. For each reply, write the exact question you asked, the date, who answered, and the applicant profile the answer assumed. Most pairs stop disagreeing on the page.
What an answer is actually worth
Normalised, the replies are still not equal, and the ranking is not subtle.
A dated written reply through the official route, quoting your own specifics — degree title, university, target intake, and your APS registration and shipping dates — is the only answer you can act on for an edge case, because it is the only one that was ever about you. The published official pages come next: authoritative for the general rule, and deliberately silent on individual cases. Everything else — a consultant's chart, a senior's experience, a confident post in a group — is useful for orientation, not for decisions.
Keep the artefact rather than the gist. "APS said it was fine" is not something you can produce later; the dated reply is. The same goes for the version of the official list you relied on — which is why checking which of your degrees the rule is about ends with keeping what you were told.
When the disagreement survives
Sometimes two answers from the official route genuinely do not line up. Do not average them, do not choose the more convenient one, and be especially careful of the one that lets you skip the test — that is the answer you want to be true, which is the worst reason to believe it.
Go back instead with a narrower question. One case, one intake, the dates attached, and a quotation of the published wording you are asking about. Say what you were told previously and by whom, so the reply addresses the conflict rather than restating the general rule you have already read. A broad question invites a broad answer, and the conflict lives in the details. If you are relying on one of the transitional exemptions, what each exemption actually asks for is the wording to quote, because each turns on a qualifier that is easy to skim.
The two mistakes do not cost the same
While the question is open, the asymmetry should shape what you do. Preparing for a test you turn out not to need costs study time and a fee. Concluding that you are outside the requirement when you are not leaves you with an APS file that cannot be completed, on a test calendar you do not control.
That second cost is not theoretical right now. On the published calendar for the first India cycle, registration closed on 15 September 2026, the test was held on 26 September 2026, and certificates are issued from 12 October 2026 — and no further test date has been published for anyone who now discovers they needed one. Waiting does not shrink the requirement either, since it applies to the Summer Semester 2027 intake and later ones, so a later intake sits further inside the rule rather than outside it.
So run the cheap hedge while the answer is pending. Keep the rest of your document package moving, since it can be verified in parallel. Run your situation through the affected check so you know which specifics your question needs to carry. And if there is a realistic chance the answer is yes, an hour with the free dMAT diagnostic tells you where you stand before the ruling arrives rather than after.
A second answer is not a coin flip. It is usually a sign that the question was not yet narrow enough to have one.
