Worked solution: A rainbow-tree proof of Ringel's conjecture (2020)
Trees can look wildly different -- a star with leaves, a long thin path, or something in between -- so no single technique handles them all. The authors show that every large enough tree falls into (at least) one of three convenient shapes: plenty of separated leaves to trim off (Case A), plenty of long thin "bare" stretches to work with (Case B), or, failing both, a tree so dominated by a few very high-degree vertices that trimming their leaves collapses it to something tiny (Case C).
Montgomery, Pokrovskiy and Sudakov (2021, Section 2, Lemma 3.5 "Case division") show that for a small constant , every sufficiently large -vertex tree falls into at least one of three cases: Case A, , where a leaf is called non-neighbouring if it is not adjacent to another leaf; Case B, , where a bare path is one whose internal vertices all have degree in ; and Case C, . The three cases overlap, but the paper only ever needs to treat a tree as belonging to Case A or B (if it does), or else Case C.
- Leaf
- A vertex of degree in a tree.
- Bare path
- A path inside all of whose internal vertices have degree exactly in (so the path is not interrupted by any branching).