CMS Advanced Sommelier common mistakes
CMS Advanced Sommelier · Understand the traps and how to avoid them.
In the deductive tasting, the candidate announces a conclusion ("this is Barolo") after noting only color and a cherry aroma, then spends the remaining minutes defending that guess instead of building evidence.
Tasting under time pressure rewards decisiveness, and candidates rehearse a 'signature' variety or region so heavily that the first familiar cue triggers recognition rather than analysis. The exam's structured-grid format makes a confident early call feel like progress.
What to do instead
Force a discipline: complete every grid field (clarity, intensity, color, development, nose intensity, aroma families, sweetness, acidity, tannin, alcohol, body, flavor intensity, finish, quality, readiness) before naming anything. Write two or three candidate conclusions with the evidence for and against each, then eliminate. Practice with a timer so the full grid becomes automatic.
The candidate describes aromas in vague hedged language ("some dark fruit, maybe a bit of earth") without committing to specific identifiers, intensity, or development level.
Fear of being marked wrong on a specific descriptor leads to safe, generic vocabulary. Candidates also lack a calibrated aroma lexicon and default to broad families because they cannot retrieve precise terms under pressure.
What to do instead
Build a personal descriptor drill: weekly blind flights where every wine must be described with at least three specific primary, secondary, or tertiary terms plus a stated intensity (light/medium/pronounced) and development call (youthful/developing/fully developed). Use a fixed vocabulary sheet and self-score specificity, not just correctness.
When asked to recommend a wine for a dish or a guest's stated preference, the candidate names a bottle and stops, offering no reasoning about weight, acidity, sweetness, tannin, or sauce interaction.
Candidates treat pairing as a memorized list of classic matches rather than a structural analysis. Under exam pressure they reach for the famous pairing and skip the explanation the examiner is actually scoring.
What to do instead
Answer in a fixed four-step structure every time: (1) identify the dominant structural element of the dish (fat, acid, heat, sweetness, umami, texture); (2) state the wine structure that answers it; (3) name a specific wine with region and style; (4) state one risk or adjustment (e.g., high-acid sauce versus high-alcohol wine). Rehearse aloud until the structure is automatic.
The candidate mishandles service mechanics under observation: decanting a mature red without a candle or light source, pouring sparkling wine down the side of a flute, or opening a bottle of aged wine without controlling sediment.
These are physical habits built in restaurants where speed matters more than ceremony. Candidates also practice tasting far more than they practice service, so the physical sequence is unpracticed and nerves degrade fine motor control.
What to do instead
Rehearse the physical sequence with a coach or video until it is rote: stand the bottle upright in advance, decant over a light source, stop at the first sediment, and pour sparkling wine down the tilted glass wall. Practice the full service sequence at least weekly with a checklist, not just before the exam.
The candidate answers a regional question by naming a grape or appellation without placing it in the correct hierarchy, climate, or soil context, and confuses neighboring appellations (e.g., treating all of Burgundy's Côte de Beaune as Pinot Noir country, or mixing up satellite appellations).
Study is often grape-first rather than map-first, so appellations float free of geography. Candidates also learn lists of names without the climatic and geological logic that explains why the boundaries fall where they do.
What to do instead
Study each region from a blank map outward: draw the boundaries, mark the rivers, aspect, and soil types, then layer in permitted varieties and styles. For every appellation, write one sentence on why it is distinct from its neighbor. Test yourself by filling in blank maps, not by rereading notes.
The candidate treats Old World and New World as a single binary and mislabels a wine's origin because it shows ripe fruit or because it shows earth, without weighing the full structural fingerprint.
Simplified heuristics ('old world equals earthy, higher acid; new world equals ripe and oaky') are taught in introductory courses and become a crutch. Modern winemaking has blurred the signals, so the heuristic fails on the exam's deliberately ambiguous wines.
What to do instead
Replace the binary with a weighted checklist: alcohol and ripeness, acid and tannin profile, oak signature, and specific regional markers (e.g., garrigue, slate, eucalyptus). Practice with deliberately ambiguous wines and justify the origin using at least three converging signals before committing.
The candidate answers spirits questions only in terms of raw material ("it's made from agave" or "it's a grain spirit") and cannot discuss distillation proof, barrel regimen, or the production step that creates the defining character.
Spirits study is often deprioritized relative to wine, so candidates memorize base ingredients and skip process. They also underestimate how much the exam tests the link between a production decision and a sensory outcome.
What to do instead
For every major spirit category, build a one-page process chain: raw material, fermentation, distillation type and proof, barrel type and time, and the sensory consequence of each step. Be able to name the step that produces the signature character (e.g., double distillation and cut points in Cognac, pot still versus column in rum, barrel char in bourbon).
The candidate describes beer only by color and broad style name, missing the fermentation family, hop character, and the specific process decisions (e.g., decoction, dry hopping, barrel aging) that define the style.
Beer is treated as a light topic and studied from taproom menus rather than from production. Candidates also confuse color with flavor intensity and cannot separate malt character from yeast character.
What to do instead
Organize beer study by fermentation family first (lager, ale, mixed and wild), then by the defining process and sensory markers. For each style, state the base malt, hop role (bittering versus aroma), yeast contribution, and one classic commercial example. Blind-taste across styles to calibrate malt versus yeast versus hop signals.
The candidate answers sake questions by treating it like wine, describing 'vintage' and 'terroir' in wine terms and failing to discuss polishing ratio, rice variety, water, or the koji and fermentation process.
Sake is unfamiliar and candidates map it onto the wine framework they know. The category's vocabulary (junmai, ginjo, polishing ratio, kimoto, yamahai) is also easy to memorize without understanding what each term does to the final product.
What to do instead
Learn sake from the production chain: rice variety and polishing ratio, water, koji, yeast, fermentation temperature, and pasteurization. Tie each classification term to a sensory consequence (e.g., lower polishing ratio and cooler fermentation toward fruity, delicate ginjo aromas). Taste across polishing ratios side by side to anchor the theory.
Across all domains, the candidate answers with a single fact and stops, rather than explaining the mechanism or the 'why' the examiner is looking for, and cannot adapt when the examiner probes with a follow-up.
Candidates prepare by memorizing discrete facts and never practice explaining them aloud. The exam's oral and analytical components reward reasoning chains, so a correct fact delivered without reasoning scores poorly and collapses under follow-up questions.
What to do instead
Practice answering every question in a fact-mechanism-consequence structure: state the fact, explain the mechanism, and give the sensory or practical consequence. Have a study partner ask 'why?' twice after every answer. Record yourself and review for answers that end too early or lack a causal chain.