Angel number data

Two tables, both counted rather than collected, both free to reuse with attribution. Angel Clock publishes them because the numbers in this subject are usually asserted and almost never shown, and because we had to compute them anyway to build the app.

1. How often each birth number occurs

Every calendar date in the period was reduced by the arithmetic the app uses — month, day and the digit sum of the year each reduced on their own, then added and reduced once more, with 11, 22 and 33 never reduced — and the results counted. The figures below are that count, not a sample and not an estimate.

Download as CSV or as JSON — both include a which_totals column listing the exact totals, which the table above leaves out for width. The JSON carries the license and the attribution line inside the file, so it stays with the numbers wherever they go.

The column that explains the imbalance

ways in is the whole explanation of the imbalance. A birth number is decided by a raw total before anything is reduced, so the question “how rare is this number” is really “how many different totals land on it”. A nine is reached from five of them; an eight from six. But 33 is reached from exactly one — a total of 33 — because anything larger is summed back down and never returns. Twenty-two has one door as well, and so does two: eleven takes 29, 38 and 47 away from it, which is why 2 sits at 4.6% while its neighbors sit near 11%.

The consequence is testable on yourself. Add the digits of your birth year and reduce them: if you get 1, no date in that year can produce a 33 at all. If you get 22, there were twenty-seven days that could. 1133 works through it in full.

2. Every minute a clock can repeat

Read straight out of the pattern tables the app runs, so this list and the shipped behavior cannot disagree. A 12-hour face reaches seven of them; a 24-hour face reaches thirteen — and that gap is why so many questions in this subject have two different correct answers, depending on a setting in the reader’s phone.

Download as CSV or as JSON, the same rows with the license attached. Angel time explains what a 12-hour face erases and why.

3. Which clock face a country gets by default

The gap in the table above is not a curiosity — it decides what a reader is ever shown. Which face a phone ships with follows the territory it was set up in, and that preference is published: Unicode records it in CLDR, the same data every operating system reads. Crossed with the table above, it answers a question this subject usually leaves open — not what a minute means, but whether the reader’s own phone can reach it at all. A territory on a 12-hour face gets fewer patterns and more encounters, because each one comes round twice.

Download as CSV or as JSON. Territories are ISO 3166-1 alpha-2; the hour preference comes from CLDR, and the pattern counts are read from the same tables the app runs. Angel time explains the split in words.

License, and how to credit it

All three datasets are published under CC BY 4.0. Use them anywhere, including commercially; the condition is attribution — say where they came from. A line like “Angel Clock, angelclock.com/data” is enough.

One half of the third table is not ours to give away. Which face a territory prefers is Unicode’s data, taken from CLDR version 49 — Copyright © 1991-2026 Unicode, Inc., used under the Unicode terms of use. What we add is the crossing: how many of the app’s patterns each face actually reaches. If you reuse that file, credit both.

What the license covers is the counted rows: the numbers, the tallies and the shares. What it does not cover is the Angel Clock name and mark, and the app’s longer readings — the extended text behind each number, which is not published here or anywhere else on this site. The one-word labels in meaning_word are included. Saying this out loud costs a paragraph and saves an argument: a license on a whole file, with nothing marked off, reads as permission for everything in it.

If you are checking our arithmetic rather than reusing it: the reduction rule is written out in full on the angel number calculator, with worked examples. Both counts above — the distribution and the ways in column — were confirmed against a second, independent implementation before publication.

Which life path number is the rarest?

33 — 195 of the 23,376 dates, one in 120. At the other end there is no real winner: 1, 3, 5, 7, 8 and 9 all land within three dates of each other, near one in 9. The ways in column above is where the gap comes from: 33 is reached from a single raw total, because anything larger is summed back down and never returns, while a nine is reached from five. “Life path number” is the older, numerological name for the figure Angel Clock calls a birth number; the calculator writes the reduction out in full.

Can I reuse these numbers in an article?

Yes. Both tables are CC BY 4.0, so you may republish, adapt and sell work based on them as long as you credit the source — “Angel Clock, angelclock.com/data”. Angel Clock would rather be cited than paraphrased, which is why the files are here as CSV rather than locked inside a page.

Where do the 23,376 dates come from?

They are every calendar date in the period covered by the table, leap days included — not a survey and not a sample of real people. Angel Clock counts dates because that is what the arithmetic acts on; how many people happen to be born on each date is a different question, and one this dataset does not claim to answer.

Where to go from here

The calculator runs the same arithmetic on your own birth date, in your browser, with nothing sent anywhere. All numbers is the readable version of the second table, and angel time is the long answer to which minutes a clock can show at all.

Last checked 10 October 2026

The tallies here are arithmetic, not a forecast: they count how often each answer occurs, not what it means for anyone. The one-word labels are a naming convention, and nothing here is medical, psychological, financial or legal advice.