How we calculate

We build precise, single-purpose instruments for statistics, science, and quantitative work. The goal is the same every time: give you a clear, honest answer fast, then get out of the way. No signup, no ads, no dark patterns. To keep that promise we hold ourselves to a few fixed rules about how results are produced and described.

How we pick a method

For any calculation there is usually more than one accepted method. When we add an instrument, we choose the method by the same order of preference every time:

1. A published standard or reference value, where one exists: for example, the IUPAC standard atomic weights for a molar-mass calculation, or the SI definitions for a unit conversion.
2. A widely cited, peer-reviewed method, when there is no single official standard: for example, the Wilson score interval for a binomial proportion, which most statisticians prefer to the plain normal approximation.
3. Plain arithmetic the mathematics agrees on, for things like balancing an equation or propagating uncertainty, where the result is not a matter of opinion once the inputs are fixed.

Where a respected alternative exists, we say so on the tool page and explain why we chose the default. We never silently average competing methods to manufacture a single tidy number.

What “estimate” actually means

Many of our results are estimates, and we use that word on purpose. An estimate is a careful, defensible value derived from the numbers you entered, not a guarantee and not an exact measurement of the world. Two things make a result an estimate:

• A model describes a general case, not your exact situation. An ideal-gas calculation assumes an ideal gas; a real gas departs from that assumption at high pressure or low temperature. The instrument names the model it applies so you know where it holds.
• Some inputs carry uncertainty. A measured quantity has a tolerance, and that uncertainty propagates into the result. Where a method defines a margin, a confidence interval, or a significant-figure limit, we show it rather than hide it behind extra decimals.

So when an instrument reports a confidence interval or a propagated uncertainty, the arithmetic is exact on the inputs you gave, and the interpretation still depends on the assumptions the tool states on its page. We surface those assumptions wherever we can, so a number can't be mistaken for a certainty.

How we cite our sources

When an instrument relies on a published value, such as an atomic weight, a physical constant, or a critical value from a statistical table, we cite it the same way every time, with three fixed parts: the publisher who actually issued the value (the standards body, the journal, the agency, never a blog that re-posted it), a direct link so you can confirm it in one click, and the retrieved date we read and recorded it. That convention is identical across every subject: a chemistry tool pins IUPAC the same way a statistics tool pins a standard reference handbook. One format, no exceptions. Here is exactly how a pinned source reads:

How often we review

An instrument is not “done” when it ships; it is maintained. We sort tools into review bands by how fast their inputs go stale, and the cadence is a published commitment:

Reference-data tools

Tools built on published values (atomic weights, physical constants, tabulated critical values) are reviewed when the issuing body, such as IUPAC or CODATA/NIST, publishes a revision. That is the only event that moves a standard value. Tools pinned to such a value carry a visible “Last reviewed” date.

Method tools

Tools built on a formula or statistical method are reviewed when we add a related instrument or when a better-established method supersedes the default. The math itself does not expire, so the trigger is a change in best practice, not the calendar.

Owner

The Eigenstate Labs editorial process owns this cadence. The maintainers stand behind the math; we do not attach a named-individual “reviewed by” byline (see below).

What we will never do

This is the rule that overrides the others. We do not invent numbers, sources, reviewers, accuracy claims, or outcomes to make a tool look more authoritative than it is. Concretely:

• No invented citations. If we can't pin a value to a real, linkable publisher, we don't state it as fact. We either expose it as an editable assumption or leave it out.
• No fake precision. We don't report a p-value to six decimals when the method's assumptions only justify two significant figures. Where a result has a known uncertainty, we show it.
• No fabricated authority. No invented reviewer names, no made-up credentials, no badge we can't back up, no ratings or testimonials we didn't receive. We deliberately do not attach a reviewer byline we cannot substantiate, because we will not manufacture a credential we don't hold. If we ever add a qualified reviewer, their real name and credentials will appear here, never before.
• No overreach. Results are framed as educational estimates to help you reason and check your work, not as professional, financial, or clinical advice. The suite-wide disclaimer says so, and individual tools add their own honest caveats where a method's limits matter.

Per-tool methodology

Every instrument carries its own “How this is calculated” section with the exact method, assumptions, and dated sources for that tool. The list below links straight to each one. Tools that already carry structured source citations show a “sources” count; tools pinned to a published value show the date they were last reviewed.

More on how we work

Everything here is meant to be checkable. If a figure looks wrong or a source has moved, that's a bug we want to fix.

• About: who builds and stands behind these instruments.
• Disclaimer: the limits of an estimate.
• Privacy: what happens to the numbers you type.