Methodology
How The Dry Standard scores and verifies drinks
A tasted glass earns a score. A production claim needs evidence. Here’s how we keep the two clear.
Understand the scoreThe Dry Standard publishes independent reviews of beverages at 0.5% ABV or less. Scores judge drink quality. Production facts are printed only when a source will stand behind them.
Facts may be researched. Scores are tasted. A product can sit in the catalog as a researched profile before Karl has poured it. A finished research page is labeled Research record: cataloged from sourced facts, not tasted or scored. A record that still lacks a production type or an alcohol figure is labeled Still being researched. Neither receives a 0–100 score, and neither appears in Best, 85+, or score filters. The reviews list groups both under Not yet tasted.
Scoring
We use a 100-point quality scale only after a Dry Standard cellar tasting. The number is not a likeness-to-ethanol score, nor a penalty for ethanol’s absence. A formulated aperitif can outscore a clumsy dealcoholized wine.
We weigh flavor, aroma, texture, finish, complexity, balance, value, and success within the claimed category. For wine, that means how well the finished drink preserves or reconstructs wine’s structure, balance, complexity, and sensory experience—not how closely it imitates grape juice or a particular alcoholic counterpart.
Body should suit the style, not simply be full. Persistence matters when desirable aromas and flavors remain coherent; a long, cloying or harsh finish is not a virtue.
Documented production methods provide context, not bonus points. We judge the finished drink through firsthand tasting, not the reputation of its removal equipment. How is alcohol removed from wine?
The quality of structure and balance contributes to the score. Likeness and structural authenticity are separate descriptive assessments, not standalone scores.
Editorial bands
These bands guide reviewers. We do not print marketing stickers on the score badge.
| Band | Guidance |
|---|---|
| 90+ | Rare. Reorder / cellar reference. |
| 85–89 | Standout in category. Shortlist material. |
| 80–84 | Clearly good. We would seek it again. |
| 75–79 | Solid everyday bottle when the style fits. |
| 70–74 | Drinkable, with a clear flaw or thinness. Recommend only with caveats. |
| 60–69 | Adequate. There is a reason to skip it. |
| Below 60 | Not recommended. |
The score is the editor’s. The prose has to follow the glass: it may not invent a story the structure and tasting notes do not support.
Facts may be researched. Scores are tasted. Research-only cellar records are not scored and do not appear here.
How tasting works
We taste products as they are sold, using Appearance → Nose → Palate → Finish → Overall / Score. Appearance is included when observed. Structure scales are category-aware: wine fields are not forced onto beer or spirits.
For wine, we assess:
- Appearance: clarity, color, hue and visual consistency. Unusual color or haze is an observation, not by itself proof of oxidation, dilution or a fault.
- Nose: aromatic intensity, clarity, varietal expression where applicable, and any observed faults.
- Palate: acidity, sweetness, tannin where relevant, body, texture, flavor intensity, integration and balance.
- Finish: persistence, flavor carry and the sweetness, dryness, bitterness, acidity or tannin that remains. Finish is a distinct part of our tasting, not just an extension of the palate.
- Overall / Score: how balance, expression, technical quality and enjoyment come together, including the effect of any observed faults.
For dealcoholized wine, we consider the effects of alcohol removal visible in the finished glass: aroma delivery and retention; acidity and tannin; body and mouthfeel; balance; finish; and overall structural integrity. We reward intelligent compensation for what ethanol normally contributes to aroma delivery, texture, body, balance and finish, not an imitation of ethanol warmth. We do not infer a production cause from a tasting impression. A softer body alone is not a fault; low flavor intensity or limited persistence may be weaknesses when they leave the wine less coherent or satisfying.
Watch for exposed acidity, thin or watery texture, stripped aroma, sweetness that masks lost structure, artificial or chemical notes, carbonation unexpected for the style, oxidative character and possible microbial spoilage indicators such as unexpected effervescence, haze or off-aromas. Record only what is observed; these indicators do not diagnose contamination. Oxidative character is not automatically a fault, and carbonation may be appropriate to the wine. Sensory oxidation is not a measure of future oxidation risk: discuss risk only when supported by sourced process, packaging or storage evidence. A tasting cannot establish microbial stability or shelf life.
Legs or tears do not influence scores. Residual sugar and viscosity-enhancing ingredients can affect them; they are not reliable measures of body or quality in nonalcoholic wine. Body and texture are assessed on the palate.
Public glance vocabulary: Flavor profile, Structure, then Appearance / Nose / Palate / Finish when recorded. The essay makes the case; it does not paste the glance fields line for line.
What “wine-like” means
Category likeness asks whether the drink behaves like its claimed category. Structural authenticity asks whether alcohol’s usual roles — body, acid, tannin, bitterness, heat, dryness, finish — are recreated without gum-and-syrup tricks doing all the work. For wine, these assessments help describe the glass; neither replaces the overall quality score. A drink can be enjoyable and not wine-like, or wine-like and flawed.
Production classification
Dealcoholized means a documented alcohol-removal step. Alternative means built as a nonalcoholic analogue from the start. Naturally low alcohol finishes ≤0.5% without a separate removal step. Hybrid is dealcoholized material plus another defining nonalcoholic ingredient, or more than one production approach. Adding the grape must back after removal stays Dealcoholized. Unclassified means we cannot place the drink in that taxonomy from citable sources. Optional construction (built wine, sparkling tea, botanical spirit alternative) sits under Alternative. It is not a peer of Dealcoholized. See What is a built wine?.
Verification is separate from type. Disclosure stance is separate again: Documented when sources support the claim, Undeclared when makers stay silent, Withheld when secrecy is on the record. ABV without a number prints as Undeclared. A known type with unnamed equipment prints Undeclared on the method line. An on-record refusal, or a proprietary claim that never names the machine, prints Withheld there instead. The note says the producer declined to name the process. Proprietary, Classified, and Refused are not method names. We do not upgrade marketing zeros into methods.
Provenance confidence
Facts such as ABV, method, and ingredients carry source metadata when known. Confidence labels (producer documented, bottle verified, retailer sourced, and others) describe the strength of evidence. Disagreements are printed, not smoothed over. The reader’s map of that order is How The Dry Standard checks a claim.
Readers can open Sources & verification on each review to inspect why we believe a claim.
Samples and independence
A sample does not guarantee a review, and it does not guarantee a favorable one. Free samples are disclosed. Commercial relationships never change the score or the tasting wording.
Prices and availability are independently observed and may change. An observed retailer listing is not an endorsement or an automatic purchase link. Reader-service exceptions and affiliate or partner links are identified where applicable. Neither commercial status nor who supplied the bottle affects scores, Best rankings, classification, verification, editorial conclusions, or coverage.
Retasting and producer claims
Products change: reformulations, vintages, ABV, methods, packages. We keep one product record so history can accumulate without overwriting prior reviews. A later recipe or a different vintage keeps its own tasting, and the earlier score stays with the formulation that was poured. Producer claims are accepted only with a citable source; label and technical sheet beat retailer copy for process claims.
Who signs
Karl Hill is the founder and editor in chief. Research tools may help assemble sources. They do not invent a spinning cone, and they do not publish.