Editorial Policy
How a formula gets from a published source onto a page here: where we look, how we test, when we update, and what happens when we get something wrong.
Where our formulas come from
We start from primary sources, not from other calculators. Copying a competitor's output propagates their mistakes, and calculator sites copy each other constantly — which is how the same rounding error ends up on twenty domains.
In practice that means: health formulas from the papers and guidelines that defined them, so BMI thresholds come from the World Health Organization and basal metabolic rate from the Mifflin-St Jeor and Harris-Benedict publications; financial formulas from standard actuarial and accounting definitions rather than from a bank's marketing page; construction conversions from published material densities and coverage rates; statistical formulas from standard texts, with the sample-versus-population distinction stated explicitly rather than assumed.
Where two credible versions of a formula exist, we implement one, name it on the page, and explain what the other one does differently. Hiding that choice would make the tool look more certain than it is.
How each calculator is tested
Every calculator is tested against three classes of input before it goes live:
- A normal case. Realistic values, checked against an independent hand calculation. If the two disagree by more than the stated rounding, it does not ship.
- A boundary case. Zero, the maximum we allow, and the point where a denominator would vanish. Divide-by-zero must produce a plain-English message, never
Infinity. - An invalid case. Text in a number field, a negative where negatives make no sense, an absurd magnitude, scientific notation. The field must explain itself rather than silently accept the value.
We also check that unit toggles convert correctly in both directions, that a share link restores the exact inputs that produced it, and that no result ever reaches the page as NaN, Infinity or undefined.
How the pages are written
Each page is written to answer the question, then to say when the answer stops being useful. The structure is fixed — how to use it, the formula with every variable defined, at least one example worked step by step, how to interpret the result, where the method breaks down, and the questions people actually ask — because a predictable structure is easier to scan than a clever one.
Three rules govern the prose. Lead with the answer, then explain. Show real numbers rather than describing arithmetic in words. And write the limitations section honestly, even when it undercuts the tool: the BMI page says plainly that BMI misclassifies athletes, because a reader who does not know that is worse off for having used it.
We do not accept guest posts, sponsored articles or paid links. No advertiser, affiliate or partner has any influence over what a page says or what a calculator returns.
How often pages are updated
Every page carries a visible review date. Tools built on fixed mathematics — percentages, area, roots, date differences — change rarely, and their review date moves when we revisit the writing rather than the formula. Tools that depend on external figures are different: marketplace fee schedules, health guidelines, standard thresholds and typical material densities all move, and when they do the page and its worked example are updated in the same pass so the two never disagree.
We aim to review every page at least annually and to update fee-dependent pages within a month of a published change we are aware of.
Corrections
If a calculator is wrong, we want to know before your accountant does. Send the calculator name, the inputs you used, the result you got and the result you expected to info@factonisam.com.
What happens next: we acknowledge the report, usually the same day. We reproduce it. If it is a genuine error we fix the formula, recalculate the worked example, and publish — typically within a day of confirming it. If the tool turns out to be right, we reply explaining why, and if the confusion came from unclear wording we rewrite the wording, because a page that is technically correct and routinely misread is still our problem.
Substantive corrections that change a published result are noted on the page. We do not quietly edit a number and move on.