About Miracles Jar

Miracles Jar takes reported miracles seriously enough to examine them carefully — which means being honest about what the evidence can and cannot establish. We are not here to debunk or to promote. We are here to weigh the documentation and say, as plainly as we can, how likely it is that a claim has no natural explanation.

Our mission

1. To record the works of God — miracles, big and small.
2. To weigh them honestly— applying the scientific method, statistical thinking, and an evidence-based approach so you can judge for yourself what's true.
3. To give everyone a place to share — people of all races, creeds, and backgrounds telling their stories and encouraging the world with good news.

Who we are

We are just a team of volunteers, writers, techies, and every-day people who have experienced the hand of God and powerful miracles in our own lives — from blind eyes being opened, to tumors disappearing, to miracle babies. We are grateful for what God has done.

Whenever we share these stories, it seems to encourage other people and lead to even greater testimonies. So we built a place where everyone can get the good news — and add their own.

Editorial standards

Independence.Submitters never influence the assessment of their own story. Where the editors have a personal stake — including our own founders’ story — the page carries a visible disclosure and is held to the same standard as everything else.
Corrections. Every probability is an estimate, and estimates get revised. When new documentation surfaces, in either direction, the assessment changes and the page shows it. If you spot an error of fact, tell us and we will check it against the sources.
Sourcing. Claims cite the documentation we actually reviewed. When a load-bearing fact rests on a single source, or on testimony alone, the evidence ledger says so plainly.

What we mean by “a miracle”

We use the word in two distinct senses — and they need different yardsticks:

Mode A — Violation

An apparent violation of natural law — something impossible under our current understanding of the physical world (a regrown limb, a terminal cancer vanishing overnight). Here the question is: is there an adequate natural explanation?

Mode B — Providence

Naturally possible, but extraordinary in timing— the fog that covered Washington's retreat; the right person at the right moment. No law is broken. Here the question is: how improbable is it that this was mere chance?

Both get an honest probability — but a Mode-A number answers “could nature do this?” while a Mode-B number answers “was this just coincidence?” We label which lens a claim uses so the figure isn't read the wrong way.

1. We gather the sources and rank them

Every source is tagged by its proximity to the event:

Primary — direct or contemporaneous records: medical files, sworn testimony, documents written at the time by witnesses.
Secondary — investigations, journalism, and scholarship that work from primary material.
Tertiary — downstream retellings: popular books, devotional accounts, websites. Useful for the story; weak as evidence.

2. We build an evidence ledger

Each consideration is logged as pointing toward an authentic miracle, toward a natural explanation, or as neutral context — and is rated weak, moderate, or strong. The ledger is shown in full on every claim so you can disagree with our weighting.

3. We publish an explicit probability

We state a single number — P(authentic), the estimated probability that no adequate natural explanation accounts for the event — paired with a plain-language verdict band and a separate confidence rating. We never publish 0% or 100%: a claim is never proven miraculous, and “no natural explanation yet” is never proven impossible, so every estimate is clamped to 1–99%.

These are subjective Bayesian estimates, not measured frequencies. They are our best read of how the documentation shifts the odds. A confident-sounding number on thin sources is dishonest, so confidence is labeled separately from the estimate itself — and both move as better evidence surfaces.

High confidence — strong primary sources, few unresolved gaps.
Medium confidence — credible but incomplete documentation.
Low confidence — thin, late, or unverifiable sourcing; the estimate is provisional.

4. We revise

Assessments are living entries. Better sources, a debunking, or a medical follow-up can move a number in either direction. If we got something wrong, we want to fix it — corrections and new sources are welcome.

5. Anyone can submit

Most miracle accounts never get written down. So you can share a story directly. New submissions post to the research queue marked “not yet validated,”and readers upvote the ones we should research and assess next. The story comes first; the assessment is added when we've done the work.

Get involved

We are all busy people who do this on the side — and we'd always welcome more help. If you have something to offer and would like to volunteer, email us at info@miraclesjar.com — or simply share a story.