A wonderful coincidence or an expected connection: why π² ≈ g

Original Article ↗ Hacker News Discussion ↗

Origin of the π² ≈ g Relationship

  • Many point out that π² ≈ 9.8 is not a “mystical” fact about Earth, but a consequence of how the meter and second were historically defined.
  • The key link is the simple pendulum: (T = 2\pi\sqrt{L/g}).
    • If you choose the unit of length so a pendulum of unit length has a 2‑second period, then (g = \pi^2) in those units.
    • Early metric proposals (seconds pendulum, toise, etc.) effectively tied the meter to this relation, so π² ≈ g in m/s² is baked into metrology, not nature.
  • Later redefinitions (Earth meridian, then speed of light/atomic time) obscured this origin but left the numerical closeness.

Units, Coincidences, and Heuristics

  • Strong debate over a common heuristic: “If a relation disappears when you change units, it’s probably a coincidence.”
    • Some argue this heuristic mostly works and that the post’s phrasing is misleading.
    • Others note this case is precisely about unit definitions, so dependence on units is the signal, not noise.
  • Distinction emphasized between:
    • Dimensionless constants lining up (often meaningful), vs.
    • Dimensionful quantities matching “nice” numbers (usually arbitrary unless unit definitions encode the relation).

Variation of g and Nature of π

  • Several note that g varies over Earth and across celestial bodies, so any numerical value of g is inherently local and conventional.
  • Side debate on whether π is “the same everywhere”:
    • One side: π is a fixed mathematical constant; non‑Euclidean circles just have different circumference/diameter ratios.
    • Other side: from within curved spaces, that ratio is not constant, which is conceptually interesting even if π (as defined analytically) is fixed.

Other Numerical and Unit “Coincidences”

  • Many share analogous curiosities:
    • c ≈ 1 foot per nanosecond; year ≈ π·10⁷ seconds; mile ↔ km via φ; insolation ≈ 1 kW/m²; Avogadro’s number × Boltzmann ≈ gas constant ~ 1 in human‑scale units.
    • Fun imperial/metric near‑equalities (yards vs meters, inches vs mm, hardware fits).
  • Consensus: such patterns range from pure coincidence to “designed” via unit choices; distinguishing them is part of good physical reasoning.

Historical Units, Metrology, and Broader Reflections

  • Discussion of Sumerian, royal cubits, Fibonacci‑like old French units, and the cognitive ergonomics of systems (base 10 vs 12, human‑scale choices).
  • Some praise the post as a delightful metrology lesson; others find it overblown or poorly explained, stressing that real physics focuses on unit‑independent structure.
  • A few use the thread to reflect on:
    • How hard defining reproducible units actually is.
    • How easily numerology creeps into physics talk.
    • Whether current AI systems can discover such historical/structural connections on their own.