Precision versus accuracy
Precision describes how finely a number is expressed; accuracy describes how close it is to the value of interest. A calculator can display 642.37 calories even when the portion and recipe are only loosely known. Extra digits preserve arithmetic, not truth. A responsible display keeps enough precision to be useful without implying that an estimated meal was measured in a laboratory.
Identify the kind of uncertainty
Food identification, reference matching, preparation, portion size, and source-data completeness each contribute uncertainty. They should not be collapsed into one vague disclaimer. Naming the uncertain input makes review actionable: a user can correct a missing sauce, choose the cooked rather than raw reference, or replace a visual portion with a known label serving.
Three different kinds of range
A portion scenario changes quantity while holding the reference food constant. A recipe scenario changes ingredients or preparation. A statistical interval describes uncertainty under a defined statistical model. These ranges answer different questions and should not be labeled interchangeably. Feast nutrition-page ranges are portion scenarios unless a page explicitly says otherwise.
Rounding policy
Feast reference pages calculate from the source value and documented portion weight before applying display rounding. Whole-number calories are generally more readable, while small nutrient amounts may need one or two decimals to avoid rounding everything to zero. The underlying calculation can retain more precision than the interface shows, but the displayed result should remain proportionate to the evidence.
Use decision-relevant precision
The right level of detail depends on the decision. A rough comparison between two visibly different portions may not need a single-number answer. Reproducing a labeled serving can use the label's units. A clinical restriction may require inputs and professional oversight beyond a consumer estimate. Do not collect extra measurements merely because the interface accepts them; collect the facts most likely to change the decision.
Communicate the dominant uncertainty
A useful note identifies the variable that could move the estimate most: unknown cooking oil, bowl depth, recipe yield, restaurant customization, or a missing nutrient field. If two plausible assumptions produce meaningfully different answers, show the scenarios. If they would not change the practical conclusion, avoid spending time polishing minor decimals.
Limits and cautions
- This resource is educational and does not provide individualized medical or nutrition treatment.
- Methods and product behavior are reviewed periodically; the displayed review date indicates the current editorial version.
Sources and further reading
- Food Labeling Guide — U.S. Food and Drug Administration Accessed 2026-08-01.
- Feast - Caloric Intelligence — Apple App Store Accessed 2026-08-01.