THE FOLD / BOSS / SUDDEN DEATH / THE ISOPERIMETRIC
THE ISOPERIMETRIC
the circle enclosing the most area for its perimeter
1 WHAT IT IS · WHAT IT DOES · FACT OR FICTION
The isoperimetric inequality answers the oldest optimization question: of all closed curves with a given perimeter, which encloses the most area? The answer — known to the ancients as ‘Dido’s problem’ but only rigorously proved in the 19th century — is the circle. For any simple closed curve of length L enclosing area A: 4πA ≤ L², with equality only for the circle. The ratio 4πA/L² (the ‘isoperimetric quotient’) is at most 1, and a regular n-gon achieves π/(n·tan(π/n)), which climbs toward 1 as the polygon rounds out into a circle.
LIT verified live: for tens of thousands of random convex polygons, 4πA/L² never exceeds 1, and the regular n-gon quotient π/(n·tan(π/n)) increases toward 1 (0.605, 0.785, 0.907, 0.977, 0.999 for n = 3, 4, 6, 12, 60) (window.__isoperimetric). FIG no framing; the polygon areas, perimeters, and quotients are computed independently in-browser.
LIT verified live: for tens of thousands of random convex polygons, 4πA/L² never exceeds 1, and the regular n-gon quotient π/(n·tan(π/n)) increases toward 1 (0.605, 0.785, 0.907, 0.977, 0.999 for n = 3, 4, 6, 12, 60) (window.__isoperimetric). FIG no framing; the polygon areas, perimeters, and quotients are computed independently in-browser.
2 HOW IT WAS WEAVED · AI + HUMAN
David (human) seated this at sudden-death — the boss round no shape can beat: for a fixed perimeter, the circle takes the maximum area and every other curve loses. AVAN (AI) built the instrument: the polygon area/perimeter, the quotient 4πA/L², and the regular-n-gon limit.
Credit as content: the classical isoperimetric problem (Dido’s problem; Steiner, Weierstrass, and others for the proof). The weave: David names the unbeatable circle; I confirm 4πA ≤ L² with the circle alone at equality.
Credit as content: the classical isoperimetric problem (Dido’s problem; Steiner, Weierstrass, and others for the proof). The weave: David names the unbeatable circle; I confirm 4πA ≤ L² with the circle alone at equality.
3 ONE DIMENSION
A polygon and the circle of the same perimeter — the circle always encloses more area.
4 TWO DIMENSIONS · INTERACTIVE
Grow a regular n-gon; the quotient 4πA/L² climbs toward 1 as it rounds into a circle.
5 THREE DIMENSIONS + AVAN’S INVERSE
The green forward object: the circle, the shape of maximal area for its perimeter.
AVAN’s addition (the inverse-companion): don’t ask which curve is biggest — ask which one is worst-case tight. The inverse of ‘maximize area for fixed perimeter’ is ‘the quotient 4πA/L² ≤ 1, hit only by the circle’. Magenta is the polygon losing area to the bound; green is the circle sitting exactly at quotient 1. The perimeter’s most efficient shape.
LIT Genuine isoperimetric inequality (Dido's problem; Steiner, Weierstrass et al. for the proof). Verified live: for ~20000 random convex polygons 4πA/L² never exceeds 1, and the regular n-gon quotient π/(n·tan(π/n)) increases toward 1 (window.__isoperimetric.ok, .maxR, .ngon).
FIG No framing; the polygon areas, perimeters, and quotients are computed independently in-browser. The AVAN inverse is honest — instead of asking which curve is biggest, ask which one is worst-case tight: the inverse of 'maximize area for fixed perimeter' is 'the quotient 4πA/L² ≤ 1, hit only by the circle'. Magenta is the polygon losing area to the bound; green is the circle sitting exactly at quotient 1. The perimeter's most efficient shape.
FIG No framing; the polygon areas, perimeters, and quotients are computed independently in-browser. The AVAN inverse is honest — instead of asking which curve is biggest, ask which one is worst-case tight: the inverse of 'maximize area for fixed perimeter' is 'the quotient 4πA/L² ≤ 1, hit only by the circle'. Magenta is the polygon losing area to the bound; green is the circle sitting exactly at quotient 1. The perimeter's most efficient shape.
◆ sealed .dlw.fold → folded to ROOT_0 · a sphere of SUDDEN DEATH · David Lee Wise (ROOT0), with AVAN