Pi is not random in the way most people mean the word. It’s a fixed, deterministic number that never changes. Every digit is locked in place, and calculating pi will always produce the same result. Yet the digits of pi behave remarkably like random numbers, passing virtually every statistical test for randomness that mathematicians have thrown at them. This paradox sits at the heart of one of the most fascinating unsolved questions in mathematics.
Deterministic but Unpredictable-Looking
Randomness, in everyday language, means you can’t predict what comes next. By that definition, pi fails immediately. Pi is the ratio of a circle’s circumference to its diameter, a constant that can be calculated to any precision using well-known formulas. There’s no dice roll involved. The 300 trillionth digit (which happens to be 5, as confirmed by a 2025 world-record computation) was always going to be 5.
But here’s what makes pi strange: even though every digit is predetermined, no one has found a shortcut to predict any particular digit without computing all the ones before it. The digits don’t repeat, they don’t follow an obvious pattern, and they don’t settle into any rhythm. In that practical sense, the digits of pi look indistinguishable from a sequence generated by flipping a ten-sided die billions of times.
What “Normal” Means in Math
Mathematicians have a precise term for what most people are really asking about: normality. A number is “simply normal” in base 10 if each digit (0 through 9) appears exactly one-tenth of the time as you look at more and more digits. A number is fully “normal” if every possible string of digits appears with the expected frequency. That means pairs like 14 or 79 each show up one-hundredth of the time, three-digit strings like 359 appear one-thousandth of the time, and so on for strings of any length.
If pi is normal, it contains every finite sequence of digits you can imagine: your phone number, your birthday written out in digits, the complete works of Shakespeare encoded as numbers. Not because pi is magical, but because that’s what it means for every possible string to appear at the expected rate in an infinite sequence.
No One Has Proven Pi Is Normal
Despite centuries of study and trillions of computed digits, mathematicians have never been able to prove that pi is normal, or even simply normal. It remains one of the major open problems in number theory. The statistical evidence is overwhelming: across 300 trillion digits, each digit from 0 to 9 appears almost exactly 10% of the time, and patterns distribute just as you’d expect from a normal number. But statistical evidence isn’t proof. Mathematicians need a rigorous logical argument that holds to infinity, and no one has found one.
This isn’t unique to pi. Proving normality turns out to be extraordinarily difficult for almost any specific number. We know that “almost all” real numbers are normal (in a precise mathematical sense, the exceptions have measure zero), but we can’t prove it for most of the famous constants we care about, including the square root of 2, the natural logarithm base e, or pi.
Patterns That Shouldn’t Surprise You (but Do)
People love finding patterns in pi, and the most famous example is the Feynman point: starting at position 762, there are six consecutive 9s in a row. This is also the first place in pi where any digit repeats four or five times consecutively, making it feel shockingly early. For a truly random sequence, the probability of a specific six-digit string showing up within the first 762 digits is roughly 0.08%. Unlikely, but not impossible, and with infinitely many digits to work with, coincidences like this are inevitable.
Other curiosities abound. The string 0123456789, all ten digits in order, first appears after about 17.4 billion decimal places. The digits 141592 (pi’s own decimal expansion, minus the leading 3) appear right at the start, which feels circular and strange but is just another coincidence in an infinite sequence. None of these patterns tell us anything deep about pi’s structure. They’re exactly what you’d expect from a number whose digits mimic randomness.
Can You Use Pi as a Random Number Generator?
In theory, you could treat pi’s digits as a source of pseudorandom numbers, and some people have done exactly this for non-critical applications. The digits pass standard randomness tests, and they’re reproducible, which can be useful for certain simulations. In practice, though, it’s a poor choice. Computing digits of pi is computationally expensive compared to standard pseudorandom number generators, which can produce statistically comparable output much faster.
For anything involving security, like cryptography, pi’s digits are entirely unsuitable. They’re publicly known and deterministic. Anyone who knows you’re pulling digits of pi starting at some position could reproduce your entire sequence. Cryptographic applications require generators whose output can’t be predicted even by someone who knows the algorithm, which is a fundamentally different property than “looking random.”
Random Behavior Without Randomness
The honest answer to “is pi random?” is that the question reveals something interesting about randomness itself. True randomness means there is no underlying rule generating the sequence. Pi has a rule; it’s completely determined by geometry and mathematics. But the output of that rule, its infinite string of digits, is so complex and patternless that it mimics randomness to a degree that no one has been able to distinguish the two. Whether that complexity holds all the way to infinity, in the rigorous sense captured by normality, remains something mathematicians believe but cannot yet prove.