Take any solid object whose surface is made of flat polygons — a cube, a pyramid, a soccer ball pattern, the cut-glass facets of a chandelier. Count its vertices (the corners where edges meet), its edges (the line segments where faces meet), and its faces (the flat polygon panels). Combine them with the simplest arithmetic: vertices minus edges plus faces. The answer, astonishingly, is always the same: two.
This is Euler’s polyhedron formula, announced by Leonhard Euler in 1750:
The cube has , , , giving . The tetrahedron has , , , giving . The icosahedron, with its twenty triangular faces, has , , , giving . Every convex polyhedron you can imagine — squat or elongated, regular or irregular — obeys it. The constancy is a hint that the formula is measuring something deeper than the specific shape, and the path it opened up became, after another century of work, the entire subject of topology.
The cube as a starting point
The most concrete way to feel the formula is to count on a familiar solid. Here is a cube with its corners marked.
The eight red dots are the vertices, the dark lines between them are the edges, and the polygons that the edges enclose are the faces. The arithmetic is uneventful — — and you can repeat it on any other polyhedron with the same outcome. The interesting question is why the answer is always two, no matter what shape you choose.
A sketch of the proof
The cleanest proof, due to Cauchy, has three steps. First, imagine puncturing one face of the polyhedron and stretching the surface flat onto a tabletop. The result is a planar graph: the vertices stay, the edges stay, and the faces become regions of the plane — with one detail. The punctured face becomes the unbounded outer region. So if the original polyhedron had faces, the flat graph has regions in total (including the outer one). This stretching does not change the count of vertices, edges, or faces.
Second, triangulate the planar graph: keep drawing diagonals across any face with more than three sides until every face is a triangle. Each diagonal you add increases by and by , leaving unchanged.
Third, peel off triangles one by one from the outside. Two kinds of triangle can be removed. A triangle that touches the outside along a single edge: removing it deletes edge and face, leaving unchanged. A triangle that touches the outside along a vertex and two edges: removing it deletes vertex, edges, and face, again leaving unchanged. Peel until just one triangle is left: , , (the triangle itself and the unbounded outside), and . Since no step in the process changed , the value must have been at the start as well. The formula is proved.
The hinge of the argument is that deformations of the surface — flattening, stretching, triangulating — leave the alternating sum untouched. The arithmetic is not really geometry; it is topology. It is reading off a property of the shape of the surface, not of the specific corners and edges you have drawn on it.
Why there are exactly five Platonic solids
One of the most beautiful consequences of Euler’s formula is a complete enumeration of the Platonic solids — convex polyhedra whose faces are all identical regular polygons and whose vertices all look identical. Suppose every face has sides and every vertex meets edges. Then counting edges two ways gives (each edge bounds two faces) and (each edge ends at two vertices). Substitute into and reorganize: . Since , this forces .
With , only five pairs satisfy that inequality, and each produces exactly one solid:
\text{Solid} & V & E & F \\ \hline \text{Tetrahedron} & 4 & 6 & 4 \\ \text{Cube} & 8 & 12 & 6 \\ \text{Octahedron} & 6 & 12 & 8 \\ \text{Dodecahedron} & 20 & 30 & 12 \\ \text{Icosahedron} & 12 & 30 & 20 \end{array}$$ Every row dutifully satisfies $V - E + F = 2$. There is no sixth row, because the inequality $\tfrac{1}{p} + \tfrac{1}{q} > \tfrac{1}{2}$ admits no further integer solutions. The fact that the universe of regular convex polyhedra is exactly five strong — known to the ancient Greeks but proved to be a *complete* list only with Euler's formula — is one of the most surprising scarcity results in all of mathematics. Plato, who attached the solids to the four classical elements (earth, air, fire, water) and the cosmos (the dodecahedron), would have appreciated the elegance of the cap. ## When the formula breaks — and what it teaches Try the formula on a solid with a hole through it — say, a polyhedral picture frame, or a polyhedral doughnut. Now $V - E + F$ comes out as $0$, not $2$. Try it on a double-holed solid (a polyhedral pretzel) and you get $-2$. In general, a closed surface with $g$ holes gives $V - E + F = 2 - 2g$. The number on the right is determined entirely by the *topology* of the surface — the question of how many holes it has — and is unaffected by the way the surface is divided into polygons. This phenomenon is so important that the value $V - E + F$ on a surface is given its own name: the **Euler characteristic**, denoted $\chi$. The sphere has $\chi = 2$, the torus has $\chi = 0$, the double torus has $\chi = -2$, and so on. It is one of the simplest topological invariants — quantities that do not change when a space is bent, stretched, or otherwise deformed without tearing. Two surfaces with different Euler characteristics cannot be continuously transformed into one another, which makes $\chi$ a quick and powerful tool for telling shapes apart at the level of topology. The historical arc here is striking. Euler set out to verify a curious coincidence about counting faces. His formula turned out to be the visible tip of an iceberg whose underwater bulk is the whole field of algebraic topology — a body of mathematics that classifies the *shapes of spaces* using arithmetic invariants, and which now underpins everything from the analysis of data clouds to the geometry of the universe itself. From counting the corners of a cube to the topology of spacetime: not bad for one short equation.Frequently asked
Did Euler really discover the formula first?
Almost. Euler announced the relation V − E + F = 2 in 1750, in a letter to Goldbach, and published it in 1758. But the French mathematician René Descartes had implicitly discovered the same identity around 1630 in an unpublished manuscript, by summing the angle deficits at the vertices of a convex polyhedron — though he never wrote the formula in its modern form. Credit is usually given to both: 'Euler's formula' for the equation, 'Descartes's theorem' for the angle-defect version.
Does the formula work for non-convex polyhedra?
Sometimes, but not always. For any polyhedron whose surface is topologically a sphere — meaning it has no holes and can be continuously deformed into a ball — the formula V − E + F = 2 still holds, even if it looks dented or non-convex. But a torus-shaped polyhedron (one with a hole through it) gives V − E + F = 0, and one with g holes gives 2 − 2g. The number on the right depends only on the shape of the surface, not the way you cut it up.
What is the Euler characteristic?
It is the topological invariant that the formula computes: χ = V − E + F, generalising to V − E + F − C + ⋯ in higher dimensions. The remarkable property is that χ depends only on the underlying surface or space, not on how you triangulate or polyhedralise it. Sphere has χ = 2, torus has χ = 0, projective plane has χ = 1, double torus has χ = −2. Two surfaces with different Euler characteristics cannot be continuously deformed into each other, making χ one of the simplest and most powerful topological invariants ever discovered.
How does Euler's formula limit the Platonic solids to five?
Combine V − E + F = 2 with the requirement that every face has p sides and every vertex meets q faces. Counting edges two ways gives pF = 2E and qV = 2E, so V = 2E/q and F = 2E/p. Substituting into V − E + F = 2 yields 1/p + 1/q = 1/2 + 1/E, which forces 1/p + 1/q > 1/2. The only integer solutions with p, q ≥ 3 are (3,3), (4,3), (3,4), (5,3), and (3,5) — the tetrahedron, cube, octahedron, dodecahedron, and icosahedron. Five solids, and no more.