This audit is for Italian medical graduates using TestBuddy for the Concorso SSM who want to read its analytics without being misled by a friendly dashboard. TestBuddy offers a large item pool, an AI assistant and a predictive score. The principal limitation is that its headline number — the predictive TB Score — is a vendor model estimate shaped by an adaptive feed, and it is easy to mistake a rising personalised average for readiness on an unseen, negatively marked paper.
What TestBuddy offers for the Concorso SSM right now
Vendor-reported unless attributed to the exam body; last checked 19 July 2026.
| Feature | What we found (vendor-reported, 19 July 2026) |
|---|---|
| Question bank | 30,000+ items across 700+ micro-topics |
| AI assistant | "Buddy" — real-time answers, personalised advice, per-question explanations, 24/7 |
| Adaptive feed | AI that "learns from your mistakes" and serves tailored questions |
| Predictive metric | "TB Score" — an estimate of the mark you would get today, updated live, broken down subject-by-subject with three focus areas |
| Explanations | Detailed commented explanations for each item |
| Analytics | Subject-by-subject statistics |
| Price | Free simulator tier; premium reported at approximately €20/month (about €239.99 for 12 months) |
| Official past questions | Not confirmed as included on the pages we reviewed — verify on testbuddy.it |
The two features that most need careful reading are the adaptive feed and the TB Score. Both are genuinely useful, and both actively shape the numbers you are trying to interpret. An audit is simply the habit of not confusing a personalised average with an exam forecast.
The exam anchor: decree, syllabus and released questions
The Concorso SSM is set annually by ministerial decree from the Ministero dell'Università e della Ricerca (MUR). It is a computer-based national exam of 140 multiple-choice questions, five options and one correct answer, in 210 minutes. Scoring is +1 correct, −0.25 wrong, 0 blank; the test is marked out of 140, with up to 7 curriculum points for a ceiling of 147, and your total sets your place in the national graduatoria that determines specialty and location. There is no official chapter-by-chapter syllabus — content spans internal medicine, general surgery, pharmacology, pathology, imaging and the major specialties — and the ministry releases past papers each year. Those released papers are your calibration standard; a third-party bank supplies the volume they cannot. Confirm the current decree before relying on any figure.
Define every metric before you trust it
| Metric | What it really measures | The trap |
|---|---|---|
| First-attempt accuracy | % correct the first time you meet an item | Your cleanest internal signal — protect it |
| Repeat accuracy | % correct on items already seen | Largely memory of those items; inflates fast |
| TB Score (predicted) | A model's estimate of today's exam mark | A modelled guess, not a ministry result — never a guarantee |
| Percentile / ranking | Position versus other TestBuddy users | Self-selected users, not the national cohort |
| Coverage | Share of the 700+ micro-topics attempted | Completion, not competence |
| Difficulty | Item difficulty, often crowd-derived | Relative to this platform's population |
| Time per item | Average seconds per question | Only meaningful under timed, no-assistance conditions |
The single metric closest to a readiness signal is first-attempt accuracy on unseen items under time, with no assistant open. Everything else is contaminated by repetition, self-selected comparison groups or model assumptions — the TB Score most of all, because it is explicitly a prediction.
Selection bias: why the feed distorts the average
An adaptive feed that "learns from your mistakes" is designed to over-sample your weak areas. That is good teaching and bad statistics: your visible accuracy is dragged down by a diet of hard items, so you may be readier than the dashboard implies — or, if you drift into comfortable micro-topics and re-see half-learned items, flattered. Either way the number is not comparable week to week, and it is certainly not comparable with a mixed, unseen, 140-item paper. The TB Score inherits all of this, because it is computed from the same personalised history. Read it as a trend line for effort allocation, never as a forecast of your rank.
Blueprint audit: attempted distribution versus the exam's real spread
The Concorso SSM has no fixed per-subject weighting, which makes a coverage audit more important, not less: without an official quota, an adaptive feed can quietly concentrate you in a handful of the 700-plus micro-topics while whole clinical areas go untouched. Do not trust the home-screen average. Read your attempts by subject and check that the broad areas the exam samples — internal medicine, surgery, pharmacology, pathology, imaging, and the major specialties — are all represented in roughly the proportion they appear in released papers. Flag any area sitting at a handful of attempts. Those are your override targets. This is the completion-versus-coverage problem; the blueprint-coverage matrix method adapts directly to an exam without a published blueprint.
The readiness test: five conditions for a credible signal
A number is a readiness signal only if all five hold: unseen items; timed at roughly 1.5 minutes each; mixed across subjects rather than filtered; no assistance (Buddy closed, no explanations mid-block); and an adequate sample — enough items that one good run does not swing it. TestBuddy can approximate this in an early timed, mixed block before the feed has personalised heavily; after that, the cleanest way to keep meeting all five conditions is a separate unseen source, plus the ministry's own released papers.
Algorithm override rules
Override the adaptive feed deliberately when it under-serves the exam's breadth:
- Low-frequency subjects you have barely attempted — force them in with subject filters.
- Image and data-interpretation items, which adaptive feeds sample unevenly.
- Pharmacology calculations and mechanisms, where accuracy decays without deliberate practice.
- Ethics and epidemiology / hygiene, easy to neglect because they feel peripheral to clinical stems.
Force these, review them, then return to adaptive mode. The feed optimises for engagement; your ranking depends on breadth.
Worked dashboard example: from analytics to next week's quotas
Suppose TestBuddy shows: TB Score suggesting a mid-range mark; internal medicine strong on 300 attempts; surgery weaker on 60; pharmacology weak on 25; imaging almost untouched; average time 70 seconds per item, comfortably inside the 90-second budget.
Read it as quotas. Internal medicine is well-sampled and strong — maintenance only. Surgery is under-sampled and weaker — schedule 60 fresh items. Pharmacology is both weak and thin — 40 items plus a targeted read and some calculation practice. Imaging is a coverage hole — a deliberate 30-item block the feed has not been giving you. Because your pace has headroom, spend some of it on the leave-or-answer decision that the −0.25 penalty rewards. What we did not do is convert the TB Score into a predicted rank — that number sets next week's workload, not your place in the graduatoria.
Worked example: a seven-day plan
TestBuddy does one job — adaptive teaching and targeted drilling; iatroX does one — unseen, timed measurement. No proprietary-algorithm claims; just sequencing.
- Day 1 — 40-item unseen, timed, mixed iatroX block. Record first-attempt accuracy by subject. Reference signal.
- Day 2 — TestBuddy adaptive block on your two weakest broad areas; review commented explanations.
- Day 3 — TestBuddy subject-filtered set forcing an under-served area (imaging, pharmacology calculations or ethics).
- Day 4 — Verify guideline-dependent corrections against official material; 20 consolidation items.
- Day 5 — Fresh iatroX unseen block on the same areas. Real gains appear on new items, not re-seen ones.
- Day 6 — One full 140-item official released paper under strict 210-minute timing; rehearse the −0.25 leave-or-answer call.
- Day 7 — Audit your attempt distribution, log misconceptions, set next week's quotas from the numbers — not from the TB Score.
Reading your results: three mistakes to avoid
Do not treat the TB Score as a promise — it is a model estimate shaped by a feed that over-samples your weak spots; use it as a trend, not a target. Do not read the percentile as a national rank — it compares you with self-selected users, not the concorso cohort. Do not let a healthy overall average hide an untouched subject — with no official blueprint, coverage holes are the commonest way careful candidates lose ranking places.
Continue, supplement, switch or stop
- Continue if your attempts span the exam's breadth, your unseen first-attempt accuracy is trending up, and the feed still serves genuinely new items.
- Supplement once you have seen most of the material and can no longer generate a clean unseen block — add an untouched source plus official released papers for measurement.
- Switch primary bank only for a measurable reason: repeated errors against current guidance, or analytics that cannot show attempts by subject.
- Stop trusting any metric that no longer changes your behaviour — a TB Score you already discount is not informing your study.
Bottom line
TestBuddy is a large, feature-rich Concorso SSM platform, and its AI assistant, commented explanations and subject analytics are useful for teaching and targeting. The discipline this audit asks for is narrow: treat the TB Score and the personalised average as descriptions of your behaviour inside an adaptive feed, not as forecasts of an unseen, negatively marked paper. Audit your coverage against the exam's real breadth, keep one clean unseen measurement outside the feed, and anchor calibration to the ministry's released papers.
Frequently asked questions
Is TestBuddy enough for the Concorso SSM on its own? Its reported 30,000-plus items and commented explanations are a strong content base, and the free tier lowers the barrier to sustained practice. What one platform cannot supply is an independent readiness signal once you have seen most of its items and once its own model is predicting your score — at that point most candidates add the ministry's released papers for calibration and an unseen source for measurement. Figures were vendor-reported on 19 July 2026 and should be reconfirmed.
Which Concorso SSM component does TestBuddy not reproduce well? The negatively marked, whole-paper endurance test under exam conditions is the hardest thing to reproduce, and — because there is no fixed syllabus and the feed personalises — even coverage across every clinical area needs deliberate checking. The content is there; the mixed, timed, breadth-complete experience and the −0.25 leave-or-answer judgement are best rehearsed on official released papers and timed mixed blocks.
How many TestBuddy questions should I complete per day for the Concorso SSM? There is no official number and volume is not the aim. A sustainable rhythm is 40–60 reviewed items a day, weighted towards under-attempted subjects, with at least one longer timed mixed block each week. Analysing why you missed an item and whether you should have left it blank matters more than raw throughput, given the negative marking.
When should I stop using TestBuddy and move to mixed mocks? Shift towards full 140-item timed papers once your attempts cover the exam's breadth, your unseen first-attempt accuracy has stabilised, and your pace leaves room to deliberate on borderline items. That is a coverage-and-stability signal, not a bank-completion or TB-Score signal. Finishing micro-topics or watching the TB Score rise is not the same as being ready.
How should I combine TestBuddy with iatroX without duplicating practice? Give each one job. TestBuddy is your adaptive teaching and drilling engine; iatroX supplies fresh, unseen, timed blocks that measure transfer. Never re-attempt an item you have already seen in the other tool — the value of the second source is that its questions are new to you. Learn on one, be measured by the other: the two-Q-bank rule.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026. Item counts, the Buddy assistant and TB Score descriptions, pricing and whether official past papers are included are vendor-reported and change without notice; reconfirm them on testbuddy.it. Disclosure: iatroX operates a competing question bank; its role here is confined to a job TestBuddy does not claim — unseen, timed transfer measurement outside the adaptive feed. Corrections are welcome via the feedback route on iatrox.com.
References: Ministero dell'Università e della Ricerca, concorso nazionale SSM and annual decree (mur.gov.it); TestBuddy SSM (testbuddy.it) and SSM simulator (testbuddy.it/simulatore/ssm); iatroX Concorso SSM bank (iatrox.com/concorso-ssm); Your Q-Bank Percentage Is Not Your Exam Score (iatrox.com).