Pi starts as a fact about circles, circumference divided by diameter. It should have stayed there. Instead it turns up in coin-flip statistics, factorial approximations, needle-dropping experiments, and the uncertainty at the bottom of quantum mechanics, places with no circle anywhere in sight.
Twelve appearances
geometry · number theory · probability · physics
geometry
C = 2πr
Circumference
The definition pi was named for.
geometry
A = πr²
Circle area
Same constant, one power higher.
geometry
V = 4/3·πr³
Sphere volume
Pi survives the jump to 3 dimensions.
number theory · 1734
Σ 1/n² = π²/6
Basel problem
Euler's shock: no circle, yet pi appears exactly.
calculus
∫e−x²dx = √π
Gaussian integral
The bell curve's area is literally √π.
probability · 1733
P ≈ 2L/(πd)
Buffon's needle
Drop needles on lined paper, measure pi.
combinatorics
n! ≈ √(2πn)(n/e)ⁿ
Stirling's approximation
Factorials quietly carry a pi.
quantum mechanics
ΔxΔp ≥ ħ/2
Heisenberg uncertainty
ħ = h/2π. Pi sets the floor on precision.
signal processing
1/√(2π) · ∫f(x)e−iωxdx
Fourier transform
The normalizing constant from panel 03's series.
electromagnetism
F = q₁q₂/4πε₀r²
Coulomb's law
Pi from a charge's field spreading over a sphere.
thermodynamics
B(ν,T) ∝ ν³/(ehν/kT−1)
Planck's law
Pi enters integrating radiation over all directions.
1748
eiπ + 1 = 0
Euler's identity
Growth, rotation, and pi in one line. See page 03.
01
A sum with no circle that lands on pi anyway
Add up the reciprocals of the square numbers, 1 + 1/4 + 1/9 + 1/16 + …, forever. Nothing here mentions a circle. Euler proved in 1734 that the sum converges to exactly π²/6, a result that stunned mathematicians who'd been stuck on it for decades.
1/1² + 1/2² + 1/3² + … = π²/6 ≈ 1.6449
Pi's appearance here comes from the same infinite series machinery as panel 03's Euler identity, not from any drawn circle. It's the first hint that pi is bigger than geometry.
02
Pi underwrites the bell curve
the area under e−x² is exactly √π, shaded
The Gaussian integral has no circle in its setup either, it's just the area under e−x². The clean proof squares the integral and rewrites it as a genuinely circular problem, integrating over a disk in polar coordinates, which is where the pi sneaks in.
∫−∞∞ e−x² dx = √π
Divide by √(2π) and you get the normal distribution's normalizing constant, the reason IQ scores, measurement error, and noise all carry a pi buried in their formulas.
03
Estimating pi by dropping needles
warming up…
Rule strips of paper a fixed distance apart, drop a needle shorter than that spacing, and count how often it crosses a line. In 1733 Buffon showed that crossing probability depends on pi, which means you can measure pi by throwing needles at a floor.
P(cross) = 2L / (πd)
The needles here are dropping and accumulating live. With enough drops the running estimate wobbles down toward 3.14159, purely from geometry and chance, no circle drawn once.