How the score works, and where it stops working
The analyser estimates fibre and plant variety from typical food-composition values, then measures them against targets. Each claim points to a reference labelled by kind of study, and editorial choices are marked as such.
Targets
The UK recommendation for adults.35 Higher intakes are linked with lower risk of heart disease, type 2 diabetes and bowel cancer in observational studies.33
For a single recipe the analyser uses a 10 g share per serving: 30 g a day split over three meals. A meal plan is measured against 30 g a day.
People reporting more than 30 different plants a week had different gut microbiomes from those reporting 10 or fewer.24 An association from one cross-sectional study, used as a prompt for variety, not a rule.
What the variety sub-score aims for in one meal and in one day. A longer meal plan adds 2.5 for each extra day, up to 30 for a week. These are editorial choices, not research thresholds.
What goes into the score
Recipes and meal plans get an overall score out of 100. It is a weighted sum of five sub-scores, each also out of 100, minus deductions. The weights are editorial: fibre, the best-evidenced part, carries the most, and the evidence behind each part is noted beside it.
- Fibre40%Fibre against the target for a serving or a day. The best human evidence: UK guidance and a large set of meta-analyses.3335
- Plant variety25%Plant points against 8 for a meal and 15 for a day, rising to 30 for a week. Rests on one cross-sectional study.24
- Fibre-type variety15%How many fermentable fibre types are present. Mechanistic: different types feed different microbes.18
- Fermented foods10%Live cultures count most. Rests on one small randomised trial.49
- Polyphenols10%Plant compounds that gut microbes transform. Mostly animal and lab studies.17
- Deductionsto −35Points off for processed meat, ultra-processed ingredients, and additives or sweeteners of concern. Concerns that rest only on mouse or lab studies count half.
Grades
The overall score sets a letter. Recipes and meal plans, and packaged-food labels, use the same bands with different names. Each band also lists the sample recipes, meal plans and labels on this site that land in it, worked out fresh each time this page opens.
Fibre types and the microbes that use them
Fibre is not one thing. The analyser tracks seven fibre types, the foods that carry them, the gut microbes that use them (directly, or by feeding on what other microbes release) and the short-chain fatty acids they mostly become.
The seven microbe groups
Plant points
How evidence is graded
Three levels
References by kind of study
Limitations
- Fibre values are typical figures for each food. Varieties, ripeness, cooking and portion size all move the real number.
- Standard fibre methods under-count some fermentable carbohydrates, such as the inulin in Jerusalem artichoke, so some plates are better than they look.
- The microbe links describe which bacteria can ferment a fibre type. Whether they do in a given person depends on the microbes that person already carries.
- Much of the additive evidence comes from mice, lab models of gut microbes or small, short human trials. The decoder says so on every slip.
- The 30-plants figure is an association from one cross-sectional comparison of self-selected volunteers, not a tested intervention.
- The analyser reads free text. Unusual names, brands and dishes may be missed; anything it cannot place is listed rather than guessed.
- Scores, microbe bars and fatty-acid estimates are simplified models, not measurements of anyone’s gut, and they are not medical or dietetic advice.
Bibliography
Every source the tools cite, grouped by kind of study. Numbers match the source marks on this page.