WorksMathematics
The Zaporozhets-Tarasov Conjecture on Mean Distances
Abstract
For a convex body let be the expected distance between two independent uniform points of , and let be the corresponding expectation for normalized surface measure on . The Zaporozhets-Tarasov conjecture asserts . We prove this conjecture in case . In addition, we give a six-vertex convex polytope in for which the reverse strict inequality holds, and obtain counterexamples in every dimension by taking products with segments.
Introduction
Let be a compact convex body. The conjecture of Zaporozhets and Tarasov asks whether
for every convex body , where are independent uniform in with respect to the normalized Lebesgue measure and are independent uniform on with respect to the normalized -dimensional Hausdorff measure. The conjecture was proposed by D. N. Zaporozhets and A. S. Tarasov in 2019 through informal communication. To the best of our knowledge, its first explicit formulation in the literature is due to Tokmachev [3] (Eq. (3) on p. 3), who attributes it to them; see also discussion before Conjecture 1.1 in the work [2].
Lotnikov proved the planar centrally symmetric case and an analogue for sufficiently high moments in arbitrary dimension [2]. Tokmachev proved the conjecture, in fact a convex-order statement, for planar bodies whose area and boundary centroids coincide and for a class of circumscribed polytopes [4]. Related extremal questions for and were studied in [1, 3]. In particular, Bonnet, Gusakova, Thäle and Zaporozhets proved the sharp inequalities
Email address: [email protected] (Eric Shen) for planar convex bodies [1], where stands for perimeter of . It is also shown that the upper bound is approached by thin rectangles converging to a doubled segment. According to Tokmachev [4, Eq. (5)], Gusakova and Zaporozhets suggested an analogous pair of inequalities for
The upper bound, with equality only for the disk, was proved by Tokmachev [3]; the complementary lower bound was established by the author after the initial submission of the present paper, see Section 5 in [8]. Combined with the upper bound for from [1], it yields another proof of the planar Zaporozhets–Tarasov conjecture.
We prove the Zaporozhets-Tarasov conjecture in the case , see Corollary 3.1. The result is mostly self-contained, in particular, it does not use the aforementioned inequalities on from [1] and is obtained as a corollary of a more general statement, see Theorem 2.1. We also prove an analogous inequality for the second moment, see Corollary 3.2. In addition to that, we show that the conjecture is false in every dimension , see Example 4.2, Figure 1, and Proposition 4.5.

Figure 1. The polytope in the rescaled coordinates . One transverse width grows with , while the other decreases.
Definition 1.1. For a probability measure on , write
This is the Gini mean difference of .
We slightly abuse this notation throughout the text, writing with being a random variable or a probability density. In either case, is understood as , where denotes the corresponding probability measure.
We use the following standard identities for the Gini mean difference; see also [6 Sec. 2.1.1].
Lemma 1.2. Let , be probability measures on with finite first moments, and let be their distribution functions. Then
Moreover, for ,
Finally, if has an atomless distribution , then
Proof. Let , be independent with law . Since
taking expectations and applying Tonelli’s theorem gives (1.1). The distribution function of is . The identity
then proves (1.2).
Finally, if is atomless, then is uniform on . By independence and symmetry,
This proves (1.3).
The main result immediately follows from a stronger directional inequality.
Theorem 1.3. Let be a convex body. Fix a unit vector . Let be the pushforward of normalized area measure on under , and let be the corresponding pushforward of normalized arclength measure on . Then
The inequality is strict if is not perpendicular to any nondegenerate segment contained in .
Plan. After rotating and rescaling, the projection of is and
where is concave and is convex. Put
The projected area density is proportional to , while the projected boundary density is proportional to
The proof has two steps. First, we find two densities and , nonincreasing and nondecreasing, respectively, such that the projected arclength density satisfies while the projected area density is proportional to the difference of the CDFs of and . Then we apply a one-dimensional theorem saying that an endpoint mixture is more dispersed than the normalized occupation measure between the endpoints. The only substantial step is the second one, to which we now turn.
The monotone-endpoint theorem
Theorem 2.1. Let be a nonincreasing probability density and a nondecreasing probability density on . Let
Put
and assume . Define
Then
Since is concave and is convex, . Thus , and is indeed a probability density.
Remark 2.2. There is also a geometric reformulation. Let and . If , then , . Consider
Since is convex and is concave, is convex. Choosing one of the two endpoints of the (uniformly distributed) vertical slice with equal probability gives the law . Schematically,
On the other hand, the length of the horizontal section of at height is
Consequently, is the second-coordinate distribution of a uniform point of , that is, choose a slice with a probability proportional to its length and then a random point in it. Explicitly,
Every nonincreasing probability density on is a mixture of uniform densities on intervals , as shown by the following Proposition. This is a one-sided form of Khintchine’s representation theorem for unimodal distributions, see [5].
Proposition 2.3. Let be a probability density on admitting a nonincreasing representative, which we also denote by . There is a random variable such that, for independent of , the variable
has density . Likewise, for every nondecreasing density , there is such that
has density .
Proof. As the law of one may take the Stieltjes mixing measure
for which
As is nonincreasing, is nonnegative and the Stieltjes integration by parts gives
We first record two consequences of Proposition 2.3.
Lemma 2.4. Let have a nonincreasing density supported on , and let have a nondecreasing one. Then
Moreover,
Proof. By Proposition 2.3, there are mutually independent and such that and . Since is uniform on and independent of ,
Consequently,
Applying the same argument to gives .
Take independent. Direct integration gives
with the second term read as zero when . Since ,
Now let be an independent copy of . Then
Using that and are independent gives
Using
we obtain the first inequality in (5). Applying the same argument to gives the second one.
Lemma 2.5. Let , , and be as in Theorem 2.1, let , , have densities , , and , respectively, and put . Then
Proof. Recall that we denote by and the CDFs corresponding to and , respectively. Integrating the CDFs gives
Since and , integration by parts yields
Now
Expanding the two cubes and using , we obtain
The first two terms have nonpositive sum by Lemma 2.4. Dividing by proves the claim.
Proof of Theorem 2.1. Let be as in Lemma 2.5. Let denote the CDF of . Since has a density, is uniform on . Hence the Gini covariance identity 1.3 and give
where the last inequality follows from Lemma 2.5.
On the other hand, since is an equal-probability mixture of laws of and , equation (2) from Lemma 1.2 gives
Now, (5) yields
By Cauchy–Schwarz, . In fact, the inequality is strict. Indeed, is continuous, , and . Thus cannot be constant almost everywhere. Subtracting the upper bound for , and using , gives
In the right-hand side of the final inequality the former summand is non-negative, as and , where and are as in Proposition 2.3 and Lemma 2.4.
The analogous result for the second moment is immediate.
Corollary 2.6. In the setting of Theorem 2.1, let be a random variable with density , and let , as before, have density . Then .
Proof. Since is an equal-probability mixture of and ,
By Lemma 2.5,
Proof of Theorem 1.3
Fix a direction that is not perpendicular to any nondegenerate segment of , normalize its projection to , and use the notation from the plan:
The functions , are absolutely continuous and their derivatives are understood almost everywhere. The functions and are respectively nonincreasing and nondecreasing, while
The normalized projected area density is
The unnormalized projected arclength boundary density is
Let
so that the normalized projected boundary density is . Define
and
Clearly,
Since is nonincreasing and is nondecreasing, it follows that is nonincreasing and is nondecreasing. Both are nonnegative. Moreover,
and hence
and similarly
Thus and are probability densities of the type appearing in Theorem 2.1. Let and be their distribution functions. Then
Consequently,
and therefore
Since , Theorem 2.1 gives
There are at most countably many directions perpendicular to nondegenerate segments of . For these exceptional directions, the non-strict inequality follows by continuity in from the inequality already proved for all other directions. Indeed, for a fixed random vector , the dominated convergence theorem implies that the function
is continuous in , and it remains to note that and are exactly of this form. This completes the proof.
Corollary 3.1. For every planar convex body ,
where are uniform in and are uniform on .
Proof. Indeed, for normalized angular measure on ,
See e.g. [7]. Integrating (9) against the two-point area and boundary distributions, and then using (4), gives the result. The inequality is strict as (4) is strict for almost every .
Corollary 3.2. For every planar convex body ,
where are uniform in and are uniform on .
Proof. The proof of Theorem 1.3 shows that, for almost every , the projected area and boundary densities are obtained from the same pair of nonincreasing and nondecreasing densities as in Theorem 2.1. Corollary 2.6 gives
Observe that
and similarly
Therefore
for almost every . For normalized angular measure on ,
Integrating the preceding directional inequality over therefore gives
A long-body observation
In this section, we construct a three-dimensional counterexample to the Zaporozhets-Tarasov conjecture, see Example 4.2. This example is then used to construct counterexamples in higher dimensions, see Proposition 4.5. The main observation, discussed in Remark 4.4, is that in dimension three the area and perimeter profiles of sections orthogonal to a fixed direction can be controlled almost independently. We exploit this freedom to make the projected Gini mean difference strictly larger for the volume measure than for the surface measure. A sufficiently large stretch in the projection direction then yields the required counterexample.
Lemma 4.1. Suppose a random point has the form , where and the support of has diameter at most . If are independent copies, then
Proof. The lower bound follows by projection onto the first coordinate. For the upper bound, use
Thus, after division by , mean distances in a long body converge to mean absolute differences of the corresponding axial coordinates.
Example 4.2. For and , define
This is the convex hull of and the four points .
Theorem 4.3. For and , one has
Proof. The area of the section at level equals . Hence the axial coordinate of a uniform interior point has density
and distribution function
Using , we obtain
Lemma 4.1 therefore gives
We next compute the axial distribution of normalized surface measure. The two faces have combined area density
while the two faces have combined area density
The end face has area . Put
For , the boundary axial distribution function is
the remaining mass is an atom at . Consequently its Gini mean difference is exactly
The diameter of the projection of onto coordinates is at most , so Lemma 4.1 gives
For the values suggested in the statement, and hence
Remark 4.4. Consider a family of convex bodies
where the functions are nonnegative, concave and Lipschitz, and positive on . The pushforward of volume measure under has density
On the other hand, the pushforward of surface measure restricted to the part of with has density
The two end sections contribute only to the surface area. Consequently, as , the normalized volume and surface measures, projected onto the -axis, converge to distributions with densities proportional to
respectively. In dimension three this is the difference between a product and a sum . Choosing one width nondecreasing and one nonincreasing can make the normalized surface density more concentrated along the long axis than the normalized volume density. In particular, choosing and gives precisely the contrasting axial distributions used in Example 4.2. In dimension two the second expression is the empty product.
Proposition 4.5. There are counterexamples in every dimension .
Proof. The case was established above. Put . We first show that grows linearly with . By Lemma 4.1 and the computations in the proof of Theorem 4.3,
whereas
Consequently,
Thus there exists such that
for all sufficiently large .
Now let , where , and put
The boundary of is the union
The two sets intersect along , which has zero -dimensional Hausdorff measure. Hence their surface measures add. The first part has surface measure
whereas the second has surface measure
Here by we denote the volume of . Elementary computations show
Observe that the surface area of normalized on the total surface area of is
Let be normalized surface measure on . It is the product of normalized surface measure on and normalized volume measure on . If are independent with law , then, similarly to the proof of Lemma 4.1, by the triangle inequality
For , the diameter of is
Replacing normalized surface measure on by changes the corresponding mean distance by at most
Indeed, normalized surface measure on is a mixture with weight on and weight on the other boundary component. In the corresponding two-point distribution, the contribution from has weight . The remaining contribution has weight , and all distances lie in . Consequently,
Choose so that
say, , then the error terms above are . On the other hand, projection onto the first three coordinates does not increase distances, so
Combining all inequalities above, we obtain
for all sufficiently large . Thus is a counterexample in dimension .
Note added
After the first version of our work was posted on arXiv, two related independent works became publicly available. Kukushkin [9] gave a proof of Theorem 2.1 and the planar Zaporozhets–Tarasov conjecture using a quite different approach. Lotnikov [10], whose work is based on his bachelor’s thesis defended in June 2026, constructed centrally symmetric counterexamples in every dimension , for every moment of the distance. These results overlap, respectively, with our Theorem 1.3 and the higher-dimensional counterexamples.
Acknowledgements
The author thanks A. Lotnikov and D. Zaporozhets for useful discussions.
Declaration of AI use
The author used ChatGPT with the GPT-5.6 Sol model during the preparation of this work. The first proof of the planar case of Zaporozhets-Tarasov conjecture, namely Theorem 1.3, was obtained by the author without the use of AI tools; it was based on KKT conditions and an exhaustive case analysis and was substantially more technical. The argument was subsequently simplified through discussions with ChatGPT. The author also used ChatGPT to search for the explicit construction in Example 4.2 from the author’s underlying idea described in Remark 4.4 (that includes generation of relevant code). The author reviewed and verified all AI-assisted output and takes full responsibility for the content of the article.
References
- [1]G. Bonnet, A. Gusakova, C. Thäle, and D. Zaporozhets, Sharp inequalities for the mean distance of random points in convex bodies, Adv. Math. 386 (2021), 107813.DOI
- [2]A. S. Lotnikov, Mean distance between random points in centrally symmetric convex bodies, Zap. Nauchn. Sem. POMI 544 (2025), 211–235.
- [3]A. S. Tokmachev, The Mean Distance Between Random Points on the Boundary of a Convex Figure, J. Math. Sci. 286 (2024), 798–806.DOI
- [4]A. S. Tokmachev, Inequalities for convex functions of random points inside and on the boundary of convex bodies, arXiv:2607.08869, 2026.arxiv.org/abs/2607.08869
- [5]A. Ya. Khintchine, On unimodal distributions, Izv. Nauchno-Issled. Inst. Mat. Mekh. Tomsk. Gos. Univ. 2 (1938), 1–7. (In Russian.)
- [6]S. Yitzhaki, Gini’s mean difference: a superior measure of variability for non-normal distributions, Metron 61 (2003), no. 2, 285–316.
- [7]R. Schneider, Convex Bodies: The Brunn–Minkowski Theory, 2nd expanded ed., Cambridge University Press, Cambridge, 2014.
- [8]E. Shen, The Zaporozhets–Tarasov Conjecture on Mean Distances, arXiv:2608.06470v3, 2026.arxiv.org/abs/2608.06470
- [9]M. Kukushkin, The Zaporozhets–Tarasov inequality for an arbitrary planar convex body, arXiv:2608.14833, 2026.arxiv.org/abs/2608.14833
- [10]A. S. Lotnikov, Mean distance between points inside and on the boundary of a convex body, arXiv:2608.14848, 2026.arxiv.org/abs/2608.14848