Why Is Every Shuffled Deck Order Unique?
Written by Viktor Jonas · Reading time about 5 minutes
An ordinary deck has 52 cards. Shuffle it properly and you get one particular order. The question is: how many different orders are possible at all?
💬 Mia asks Grandpa Theo
52 times 51 times 50 and so on — surely that can't give thaaat many possibilities?
It does, and that is exactly the trick with factorials — they explode incredibly fast. For the first card you have 52 options, for the second only 51 (one is already taken), for the third 50, and so on down to the last card.
Okay, show me with fewer cards so I can follow it.
Gladly. With 3 cards there are 3! = 6 orders. With 4 cards already 4! = 24. With 5 cards 5! = 120. You can see it does not grow slowly but explosively — and with 52 cards you end up with a number 68 digits long.
52! ≈ 8.07 × 10⁶⁷ — that is an 8 followed by 67 digits (more precisely: 80,658,175,170,943,878,571,660,636,856,403,766,975,289,505,440,883,277,824,000,000,000,000).
For comparison: the universe is about 4.35 × 10¹⁷ seconds old. Even if every one of the roughly 8 billion people on Earth had shuffled a fresh deck every single second since the Big Bang, that would only add up to some 3.5 × 10²⁷ orders — still 40 powers of ten short of 52!
So the next time you shuffle a deck properly, the order you create almost certainly holds the record of never having existed before in human history — and it will probably never turn up again.
Factorial (n!) — the product of all the numbers from 1 to n.
Permutation — one possible order.
Combination — a selection where the order does not matter.
Laplace probability — favourable cases divided by possible cases.
How many orders there are
52! ≈ 8.07 · 1067
for comparison:
atoms on Earth ≈ 1050
seconds since the Big Bang ≈ 4 · 1017
The number is so large that any well-shuffled order has, to all intents and purposes, never existed before — not in the entire history of card playing. That is not a figure of speech but a plain consequence of the scale.
| n | n! |
|---|---|
| 5 | 120 |
| 10 | 3,628,800 |
| 20 | 2.4 · 1018 |
| 52 | 8.07 · 1067 |
How many shuffles does it actually take?
All those possibilities are worth nothing if the shuffling is poor. That is exactly what the mathematicians Dave Bayer and Persi Diaconis looked into in 1992, and their result became famous. For the riffle shuffle — splitting the deck and letting the halves fall into each other — about seven rounds are needed before a 52-card deck is effectively random. After three or four shuffles you can still detect statistically how the cards were arranged beforehand, which is too little for a fair game.
The way it happens is remarkable. The order does not fade away evenly but almost all at once. Up to the fifth shuffle a great deal of structure survives; between the fifth and the seventh it collapses; and after that, further shuffling barely changes anything. Specialists call this behaviour a cut-off. Watch a casino dealer and you will see the rule in action: the repeated shuffling is not a habit, it is there because anything less has been shown not to be enough.
The reason for the explosion lies in the multiplication. Add a single card to a deck of n and the total number of orders does not grow by a little — it is multiplied by (n + 1). Going from 51 to 52 cards multiplies the possibilities by fifty-two in one single step. This kind of growth even outpaces exponential growth: with 2n each step adds the same factor of 2, while with n! the factor itself keeps getting larger. That is why 52! is so much bigger than 252, even though both expressions look equally harmless.
In practice this means even a 32-card deck has around 2.6 · 1035 orders, and a 20-card deck still manages 2.4 quintillion. You do not need a full deck for the number to defeat the imagination — twenty cards will do.
Common mistakes
- Counting the order where it does not matter
- When drawing a hand the order usually makes no difference — then divide by the factorial of the hand size.
- Confusing drawing with and without replacement
- Cards are not put back, so the denominator drops at every step.
- Adding probabilities instead of multiplying
- For consecutive conditions you multiply.
- "Very unlikely means impossible."
- Some order always comes out.
How likely is it to draw two aces in a row from a 52-card deck without replacement?
Show solution
First ace: 4 out of 52 = 1/13.
Second ace: only 3 aces left among 51 cards = 3/51 = 1/17.
P = 1/13 · 1/17 = 1/221 ≈ 0.45 %
With replacement it would be 4/52 · 4/52 = 1/169 ≈ 0.59 % — noticeably
more. The difference arises because the first card drawn changes the starting point
for the second. That is exactly what makes card games mathematically more interesting
than dice games.
Birthday paradox · Monty Hall problem · Exponential growth · Prime numbers
🎥 Explained in under two minutes — as a Short on our YouTube channel.
Watch the Short