An inequality for anti-self-polar polytopes

This paper proves a 1989 conjecture by Katz regarding an inequality for the f-vectors of anti-self-polar polytopes by utilizing Kalai's combinatorial inequality derived from Whiteley's results, offering a more accessible alternative to previous proofs involving complex algebraic geometry.

The Gist

Imagine you are an architect designing a perfect, multi-sided gemstone (a polytope) that floats inside a giant, invisible glass sphere.

This paper is about a very special, somewhat magical type of gemstone called an anti-self-polar polytope. Here is the simple breakdown of what the author, Mikhail Katz, discovered about these shapes.

1. The "Magic Mirror" Shape

Most shapes are just shapes. But an anti-self-polar shape has a weird, mirror-like property.

• Imagine you take your gemstone and look at its reflection in the glass sphere.
• Usually, the reflection looks different.
• But for this special shape, if you flip the reflection upside down (multiply by -1) and shrink or stretch it slightly, it fits perfectly back onto the original shape.
• It's like a shape that knows its own reflection so well that they are essentially the same object, just turned inside out.

2. The "Furthest Friends" Problem

The author is interested in the vertices (the corners) of this shape. Specifically, he wants to know: How many pairs of corners are as far apart from each other as possible?

Think of the corners as people at a party.

• Some people are standing right next to each other.
• Some are across the room.
• The "diameter graph" is a map of who is standing at the maximum possible distance from whom.

The paper asks: If you have a certain number of people (corners) at this special party, what is the minimum number of "farthest-apart" pairs you can have?

3. The Big Discovery (The Inequality)

Katz proves a rule that was guessed back in 1989. The rule says:

If you have $N$ corners in this 4-dimensional magical shape, you are guaranteed to have at least $3N - 5$ pairs of corners that are as far apart as possible.

The Analogy:
Imagine you have 10 people at this party.

• The rule says: No matter how you arrange them (as long as they form this special shape), there must be at least $3(10) - 5 = 25$ pairs of people standing at the maximum distance from each other.
• It's a guarantee. You can't arrange the shape to have fewer "farthest friends" than this number.

4. How Did He Prove It? (The Detective Work)

Katz didn't just guess; he used a clever mathematical trick involving "counting faces."

• The Faces: A 4D shape is made of 3D "slices" (like how a 3D cube is made of 2D square faces).
• The Counting Game: He looked at the shapes of these slices. Are they triangles? Squares? Pentagons?
• The "Gap" (Kalai's Inequality): He used a tool invented by a mathematician named Gil Kalai. Think of Kalai's tool as a balance scale. On one side, you put the number of corners. On the other, you put the number of edges and the shapes of the faces.
• The Result: The math showed that if you try to make the shape "too efficient" (too few farthest pairs), the balance scale tips over, and the shape breaks the rules of geometry. Therefore, the shape must have that many farthest pairs to stay stable.

5. Why Does This Matter?

• Solving a Mystery: This confirms a guess Katz made 35 years ago.
• Simplicity vs. Complexity: The author notes that other famous mathematicians (Stanley and Karu) could have proven this using very heavy, difficult "algebraic geometry" (like using a sledgehammer to crack a nut). Katz used a simpler, more "combinatorial" approach (counting and logic), which is like using a precise screwdriver.
• Computer Checks: A researcher named Qingsong Wang used a computer to generate hundreds of these shapes. Every single one followed the rule perfectly, often hitting the exact minimum number predicted by the formula.

The Bottom Line

This paper is a victory for logic and counting. It proves that in the strange, high-dimensional world of these "mirror-image" shapes, there is a strict limit to how few "long-distance connections" you can have. If you have a certain number of corners, the universe forces you to have a specific number of "farthest-apart" pairs. It's a beautiful example of how geometry imposes order, even in the most abstract shapes.

Technical Summary

1. Problem Statement

The paper addresses a conjecture regarding the combinatorial properties of anti-self-polar polytopes, specifically in four dimensions ($\mathbb{R}^4$).

• Definition: An anti-self-polar polytope $P$ is a polytope inscribed in the unit sphere such that its polar $P^*$ (with respect to the unit sphere) satisfies the relation $P^* = -cP$ for some constant $c > 0$.
• Context: These polytopes were previously investigated by Lovász (1983) and Katz (1989). In 1989, Katz proposed that 4-dimensional anti-self-polar polytopes could serve as potential counterexamples to Borsuk's conjecture in $\mathbb{R}^4$. While no counterexamples have been found in this dimension, the structural properties of these polytopes remain of significant interest.
• The Conjecture: Katz (1989) conjectured a specific lower bound on the number of edges in the diameter graph $G(P)$ of a 4-dimensional anti-self-polar polytope. The diameter graph consists of vertices of $P$ connected by edges if the pair of vertices is at maximal distance (the diameter of the polytope).

2. Methodology

The author employs a purely combinatorial approach to prove the conjecture, avoiding the complex algebraic geometry typically associated with such problems.

• Key Tools:

  • Kalai's Combinatorial Inequality: The proof relies on a result by Gil Kalai (1987, 1994) concerning 4-dimensional polytopes. This inequality relates the number of $j$-gons in the facets of the polytope to its $f$-vector components ($f_0, f_1, \dots$).
  • Whiteley's Result: Kalai's inequality is itself based on a result by Whiteley (1984).
  • Euler's Formula: Applied to the facets of the polytope to relate face counts.
  • Symmetry Properties: Utilizing the specific property of anti-self-polar polytopes where the number of vertices equals the number of facets ($f_0 = f_3$).

• Alternative Context: The paper notes that the same inequality can be derived from the work of Stanley (1987) and Karu (2004) using the Hard Lefschetz theorem for intersection homology. However, the author explicitly chooses the combinatorial path to avoid the "difficult algebraic geometry" involved in those methods.

3. Key Contributions and Proof Logic

The central contribution is the rigorous proof of the conjectured lower bound on the edges of the diameter graph.

• Theorem 1: For a 4-dimensional anti-self-polar polytope $P$, the number of edges $e(G(P))$ in its diameter graph satisfies:
  $$e(G(P)) \geq 3f_0(P) - 5$$
  where $f_0(P)$ is the number of vertices.

• Proof Derivation:

  1. Relation to $f$-vectors: The author establishes that the number of pairs of vertices at maximal distance corresponds to the number of vertex-facet incidences at the "opposite" level, denoted $f_{03}(P)$. Specifically, $f_{03}(P) = 2e(G(P))$.
  2. Target Inequality: The theorem is equivalent to proving $f_{03}(P) \geq 6f_0(P) - 10$.
  3. Application of Kalai's Bound: Using Kalai's inequality ($a_4 + 2a_5 + \dots \geq 4f_0 - f_1 - 10$) and summing over all facets, the author derives a general lower bound for any 4-polytope:
     $$f_{03}(P) \geq 3f_3(P) + 3f_0(P) - 10$$
  4. Specialization: For anti-self-polar polytopes, the symmetry condition $f_0 = f_3$ holds. Substituting this into the general bound yields:
     $$f_{03}(P) \geq 3f_0(P) + 3f_0(P) - 10 = 6f_0(P) - 10$$
  5. Conclusion: Dividing by 2 confirms the bound on the diameter graph edges: $e(G(P)) \geq 3f_0(P) - 5$.

• Secondary Result: The proof demonstrates that the "gap" term $g_2(P)$ (defined in Kalai's work) satisfies $g_2(P) = g_2(P^*)$ for every 4-polytope, highlighting a duality in the combinatorial structure.

4. Results and Empirical Verification

• Theoretical Result: The inequality is proven to hold for all 4-dimensional anti-self-polar polytopes.
• Empirical Evidence: The paper cites calculations by Qingsong Wang, who generated hundreds of examples of anti-self-polar polytopes with spherical diameters in the range $\arccos(-1/4) < d < \arccos(-1/3)$ using Python and downward gradient flow.
• Observation: In all generated examples, the inequality holds as an equality (the boundary case), suggesting the bound is tight for these specific configurations.

5. Significance

• Resolution of a Long-standing Conjecture: The paper settles a combinatorial conjecture proposed by Katz in 1989, which had remained open for decades.
• Methodological Clarity: By utilizing Kalai's combinatorial inequality, the author provides a proof that is accessible to combinatorialists and discrete geometers, bypassing the heavy machinery of algebraic geometry required by Stanley and Karu.
• Insight into Borsuk's Conjecture: While the paper does not produce a counterexample to Borsuk's conjecture in $\mathbb{R}^4$, it clarifies the structural constraints of the specific class of polytopes (anti-self-polar) that were hypothesized to be counterexamples. The tightness of the bound in empirical examples suggests these polytopes are highly structured.
• Duality Properties: The result reinforces the deep duality properties of 4-polytopes, specifically showing that the combinatorial invariant $g_2$ is preserved under polarity.