The 1/3–2/3 Conjecture for finite posets

From papers

Let PP be a finite poset on nn elements. A linear extension is an order-preserving bijection bb:P[n]bb:P\to[n], and for x,yPx,y\in P define

δP(x,y)={linear extensions λ:P[n] such that λ(x)>λ(y)}{linear extensions of P}.\delta_P(x,y)=\frac{|\{\text{linear extensions }\lambda:P\to[n]\text{ such that }\lambda(x)>\lambda(y)\}|}{|\{\text{linear extensions of }P\}|}.

The balance constant of PP is

b(P)=maxx,yPmin(δP(x,y),1δP(x,y)).b(P)=\max_{x,y\in P}\min(\delta_P(x,y),1-\delta_P(x,y)).

The 1/3–2/3 Conjecture. For any finite poset PP which is not a total order, b(P)13b(P)\geq\frac{1}{3}.

This is a longstanding open problem in the theory of posets, with the bound proved for several special classes, including width-two posets. The conjecture asks for a universal information-theoretic lower bound on how balanced some pair of incomparable elements must be.

Progress summary

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

Sources & referencesView supporting material

Primary source

Christian Gaetz and Yibo Gao, “Balance constants for Coxeter groups”, arXiv:2005.09719 (2023).

Solutions 0

No solutions have been posted yet.