Forest families eventually k-nice

Question 1.1 · arXiv:1708.07369

arXiv Question high confidence— first stated 2018-04-16

Authors: Ron Aharoni, Noga Alon, Michal Amir, Penny Haxell, Dan Hefetz, Zilin Jiang, Gal Kronenberg, Alon Naor

Status open high confidence

Question 1.1 from arXiv:1708.07369 asks whether every finite family of graphs containing at least one forest is k-nice for all sufficiently large k. The source paper proves this for families containing a forest with at most 2 edges and for each of the three two-graph families of connected 3-edge graphs. A targeted search across arXiv and the published literature found no follow-up paper resolving the full conjecture, and the arXiv full-text search for 'Ramsey-nice k-nice forest' returned zero results beyond the original paper.

Reviewer notes. No follow-up resolving the conjecture was found. The arXiv full-text search returned zero results for the distinctive terminology 'Ramsey-nice k-nice forest'. The paper itself was published in European Journal of Combinatorics 72 (2018) 29–44. Question 1.1 remains open as of 2026-05-15.

Auto-reviewed 2026-05-15 with claude-sonnet-4-6 (web search enabled).

Question. Is it true that for any finite family of graphs $\mathcal{F}$ that contains at least one forest, there exists a constant $k_0 = k_0(\mathcal{F})$ such that $\mathcal{F}$ is $k$-nice for all $k \geq k_0$?

Context

It is easy to see that if $\mathcal{F}$ contains no forest then it is not $k$-nice for any $k$. The question asks whether a converse holds: every family containing at least one forest is eventually $k$-nice. The paper proves this for all families containing a graph with at most 2 edges and for several two-graph families of 3-edge connected graphs.

Source paper

Ramsey-nice families of graphs
Ron Aharoni, Noga Alon, Michal Amir, Penny Haxell, Dan Hefetz, Zilin Jiang, Gal Kronenberg, Alon Naor · 2018-04-16
https://arxiv.org/abs/1708.07369 PDF source