Bowler–Emde–Gut question on Maker–Breaker games on infinite graphs

Let B=Kℵ0B=K_{\aleph_0}, and let G1,…,Gk⊆BG_1,\dots,G_k\subseteq B be finitely many infinite subgraphs. In the color-preserving Maker–Breaker game on BB, Maker and Breaker alternately claim previously unclaimed edges of BB. Determine, for arbitrary finite kk and arbitrary choices of G1,…,GkG_1,\dots,G_k, which player has a winning strategy, where Maker wins if the edges claimed by Maker contain a copy K≅Kℵ0K\cong K_{\aleph_0} such that KK contains infinitely many edges of every GiG_i, that is, ∣E(K)∩E(Gi)∣=ℵ0\lvert E(K)\cap E(G_i)\rvert=\aleph_0 for every i∈{1,…,k}i\in\{1,\dots,k\}. More broadly, determine the winner in the corresponding partially pattern-preserving games, in which Maker must claim a copy K≅Kℵ0K\cong K_{\aleph_0} such that Gi∩KG_i\cap K is isomorphic to a subgraph of GiG_i for every ii.

References

Primary source

arXiv

Progress summary

Refreshed
Claimed progress

A new paper settles the one-pattern case and adds partial results for more patterns, but the full infinite-graph question remains open.

The Bowler–Emde–Gut question concerns which player can force prescribed colored structures in Maker–Breaker games on infinite graphs. The latest work settles the single-pattern case while leaving the general problem open.

Known results

  • Bowler, Emde, and Gut characterized the vertex-coloring game for finitely many colors, while showing limitations when infinitely many colors are allowed.
  • Related games on uncountable infinite graphs have outcomes depending on set-theoretic assumptions, including CH and MA.

August 24, 2026 partial characterization

A new paper proves Maker and Breaker conditions for k>1k>1 and completely characterizes the k=1k=1 color-preserving game. It therefore resolves the single-pattern case, but gives no complete characterization for all k>1k>1 or all pattern-preserving games; this reported advance is unverified.

Current status (as of August 2026): The k=1k=1 color-preserving case is claimed characterized, and conditions are claimed for k>1k>1, but the broader question remains open and the new results are unverified.

Sources

Solutions 0

No solutions have been posted yet.