We publish anonymized statistics from our AI tarot reading platform. Open dataset, CC BY-SA 4.0 license, methodology documented.
Current snapshot
Snapshot version: 2026-Q2 (preview) Sample size: 1,370 readings from ~750 unique participants (69 registered users + ~680 anonymous guest IPs) Time window: 2026-01-01 to 2026-05-02 Languages: EN (90.7%), PL (3.6%), PT (2.9%), FR (1.2%), ES (0.9%), DE (0.4%), IT (0.2%) Last updated: 2026-05-06
What's available now
The aggregate statistics already published on this site:
- Most drawn cards (top 15) — Knight of Wands #1 with 78 draws
- Major:Minor distribution — 28.4 / 71.6 (statistically random)
- Day-of-week patterns — Tuesday +37% above baseline
- Question category breakdown — 28.1% future-oriented, 13.4% love
- Spread popularity — 78.5% past-present-future spread
What's coming Q3 2026
The full anonymized dataset for download:
- CSV format — one row per reading, anonymized user IDs
- JSON-LD format — Schema.org Dataset markup compatible
- Per-card statistical breakdown — frequencies, position distributions, co-occurrences
We're running an anonymization audit before public release. Specifically: ensuring that the combination of (language + spread_type + week) doesn't allow re-identification of any individual user. Audit is expected to complete by 2026-08-01.
Citation
If you use any of our published statistics in research, journalism, or analysis:
aimag.me Tarot Reading Dataset (n=1,370 readings, ~750 participants). Collected 2026-01-01 to 2026-05-02. Anonymized open dataset. Available at https://aimag.me/research.
BibTeX:
@misc{aimag2026tarot,
title={aimag.me Tarot Reading Dataset},
author={Fiedoruk, Tomek},
year={2026},
url={https://aimag.me/research},
note={n=1,370 readings from ~750 participants (69 registered + ~680 guest IPs), anonymized open data, CC BY-SA 4.0}
}License
Creative Commons Attribution-ShareAlike 4.0 International (CC BY-SA 4.0).
You may:
- Copy and redistribute the data in any medium or format
- Use it commercially
- Build derivative works
Under these terms:
- Attribution required (cite as above)
- Share-alike: derivative datasets must use the same license
Why we publish this
Most "tarot statistics" you find online are made up. We collected real data because we run the tool that generated it. Publishing the anonymized aggregate is the honest move — both for the few researchers and journalists who actually want it, and for skeptics who want to verify our claims.
We have a financial interest in users finding tarot useful (we operate the aimag.me reading tool). The dataset undermines as many mystical claims as it supports. We publish both findings regardless.
Questions
For dataset questions, replication queries, or research collaboration: [email protected].
For methodology details: /research/methodology.