Editorial policy

What a page has to contain

Every calculator page carries four things: the formula in plain English, a worked example using the page's own default numbers, a statement of what the estimate cannot tell you, and a list of the published sources behind the thresholds. A page that is missing any of those is not finished, whatever else it says.

• The formula, written out with the units it expects.
• A worked example a reader can reproduce on paper.
• The limits: who the estimate does not fit and how far off it usually runs.
• Sources, linked, with the threshold each one supports.

Where the evidence comes from

Priority order: peer-reviewed literature and national guidance first - the World Health Organization, NICE, the ACSM, the NIH and the Compendium of Physical Activities - then professional body statements, then widely replicated clinical practice. Marketing pages, forum posts and other calculator sites are never used as evidence for a threshold, although they are sometimes read to see which questions people actually ask.

When two credible sources disagree, we show both numbers and say which one the calculator uses and why. The Harris-Benedict comparison on the BMR page is the clearest example: the older equation is printed beside the current one instead of being quietly discarded.

Testing before publishing

The maths lives in one file and is covered by a test suite with hand-computed values. Each formula must return the expected figure for a normal case, a boundary case such as zero or an extreme input, and a junk-input case that must not produce an error or a blank result.

Before a build is published, an automated audit checks every page for a unique title and description, a breadcrumb trail, working internal links, valid structured data, and that the result numbers are present in the HTML even with JavaScript switched off.

Review dates, changes and corrections

Pages carry a review date. A change to a formula, a threshold or a source is dated on the page that changed, and the structured data on the page carries the same date so search engines and readers see the same version.

Correction reports go to hello@calc.idailycheck.xyz. Errors that produce a wrong number are treated as urgent and corrected first; wording and clarity fixes follow. If a correction changes a figure a reader may have acted on, the note stays visible on the page rather than being quietly edited away.

How sources are ranked

Sources are graded before they are used. A primary study or a national guideline is grade A. Official documentation from a health body, such as WHO, NICE, the CDC or the NHS, is grade B. A review article, textbook or reputable explainer is grade C. A forum post, a coach's blog or our own inference is grade D.

Any claim about a threshold, a formula or a risk band has to rest on grade A or B evidence. Grade C is used only for background and context. Grade D never appears as a stated fact; if a topic cannot be supported above grade C, the page either says so explicitly or does not cover it.

When two reputable sources disagree, both positions are described and the disagreement is named, rather than picking the more exciting number. Where a guideline is region-specific, the page names the region instead of presenting one country's cut-off as universal.

No clinicians on staff, and what that means

BodyCalc is written and maintained by a small technical team, not by doctors or registered dietitians. Every page therefore stays inside what published guidance can support: population equations, screening thresholds and the documented limits of each method.

That is a deliberate boundary rather than a gap to be filled with confident-sounding copy. Nothing here interprets a result for an individual, recommends a treatment, or covers a diagnosed condition. Where a page approaches a clinical line, it stops and points to a clinician.

If you need individual advice, a registered dietitian or your doctor can weigh the things a calculator cannot see: medical history, medication, blood results, eating behaviour and what you can realistically sustain.

Independence, funding and how AI is used

The site carries no advertising, no sponsored content, no affiliate links and no paid placement. Nothing on a page is there because someone paid for it, and no company influences which formulas are used or what the results say.

Drafting is sometimes assisted by language tools, but the arithmetic, the thresholds and the sources are checked by a person against the primary documents before publication, and the calculators are validated against hand-computed values in an automated test suite. Anything that cannot be traced to a named source does not ship.

When a page fails that standard, the fix is published on the page with a date rather than applied quietly, and the correction address is on the contact page.

How a page is researched and written

Work starts from the question a reader actually types, not from a keyword list. The first step is to find out what published guidance says about it and whether a formula exists at all. If the answer depends on a threshold, that threshold has to come from a named body.

The calculator is then built and tested before the words are written, so the article describes what the tool actually does rather than what it was supposed to do. Every number quoted in the prose is checked against the tool's own output.

Finally the page is read as a whole against a checklist: one clear heading order, the formula visible, sources listed, no claim without attribution, and no sentence that could be read as individual medical advice.

When a source is withdrawn or superseded

Guidance changes. A guideline body issues a new version, a validation study finds an equation drifts in a subgroup, or a paper is corrected. When that happens to a source this site depends on, the affected pages are re-checked rather than left to age.

The order of work follows risk. A change to a threshold that a calculator uses is handled first, because it alters the numbers people act on. A change to background context can wait for the next review cycle, but it is still recorded.

Pages that depend on a superseded source are never silently patched. The review date moves, and where the change would alter a result someone may have saved, a dated note on the page says what changed and why.

Review cycle

Calculators and guides are reviewed on a rolling schedule rather than a fixed annual sweep. Anything that cites a guideline is checked when that guideline is next revised, and the highest-traffic pages are checked most often.

A review can be short. If the sources still say what the page says, only the date changes. The value of the cycle is that someone has to look at the page and decide, rather than assuming it is still right because nothing has broken.