Quick Answer: Computers use the Fisher-Yates shuffle algorithm. It goes through a list from end to beginning, randomly swapping each item with one that comes before it. It guarantees that every possible permutation of the deck is equally likely.
Why Naive Shuffles Fail
Early programmers often wrote bad shuffle code, creating a bias where certain cards ended up at the bottom of the deck. If a hacker realizes the shuffling algorithm is biased, they can statistically tilt their bets and drain an online casino. A true shuffle must be mathematically perfect.
The Fisher-Yates Logic
Imagine 52 cards on a table. The computer picks a random number from 1 to 52. Let's say it picks 14. It takes the 14th card and puts it at the very bottom (position 52). Next, it picks a random number from 1 to 51. It takes that card and puts it at position 51. It repeats this until the entire deck is jumbled. It is mathematically elegant and fully unbiased.
The Importance of Good Entropy
The Fisher-Yates shuffle is only as strong as the random number generator fueling it. If the computer uses a weak `Math.random()`, the deck is predictable. This is why PickRandom uses Cryptographically Secure Randomness (Web Crypto API) to fuel its Fisher-Yates shuffle variations.