About

A hobby project about
the nature of randomness.

In April 2025, I started wondering: if I shuffled a deck of cards continuously, how long before I saw the same shuffle twice? The answer, it turns out, is longer than the age of the universe — by many orders of magnitude.

So I wrote a script. It runs continuously, recording every shuffle to a database. Since April 27, 2025, it has run 11 sessions across ~142 days of continuous shuffling, producing 12,300,339 unique arrangements. Not one has ever repeated.

That's not surprising. A standard deck of 52 cards can be arranged in 8.07 × 10⁶⁷ different ways — a number so large it makes the number of atoms on Earth look small.

The Algorithm

Shuffles are simulated using the Gilbert-Shannon-Reeds model — the standard mathematical description of a riffle shuffle. Each shuffle cuts the deck asymmetrically near the center, then interleaves the two halves with drops weighted by remaining card count. 75% of shuffles end with a final cut.

What's Next

Eventually: sell rolls. For a small amount, get 1,000 shuffles checked against the full archive. The odds of a match are astronomically small — but that's the point. Every roll is provably fair, provably rare, and provably unique.