The variance bound conjecture for fixed-break record counts

Let KiK_i be the number of remaining records broken at time ii, let Ii,k:=1(Ki=k)I_{i,k}:={\bf 1}(K_i=k), and define

Rn,k:=i=1nIi,k,σn,k2:=VarRn,k.R_{n,k}:=\sum_{i=1}^n I_{i,k},\qquad \sigma_{n,k}^2:=\operatorname{Var}R_{n,k}.

Variance bound conjecture. For each fixed k0k\geq0,

σn,k2=O((logn)2).\sigma_{n,k}^2=O\left((\log n)^2\right).

The paper explains that this bound would yield the observations-time strong law and therefore the Glivenko–Cantelli conjecture; it is proposed because the analogous bound is known for the total record count, but the required calculation is left open.

Sources & referencesView supporting material

Primary source

James Allen Fill, “Breaking Bivariate Records”, arXiv:1901.08232 (2019).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.