{"id":"arxiv/2610.37920","url":"https://arxiv.org/abs/2610.37920","tags":["math.CO"],"score":0,"title":"High-chromatic tournament parts without a complete pair of cyclic triangles","domain":"math.CO","source":"arxiv","authors":["Guantao Chen","Hein van der Holst","Shengze Wang"],"reviews":0,"subject":"math.CO","summary":"Nguyen, Scott, and Seymour conjectured that any two disjoint subtournaments of sufficiently large chromatic number contain cyclic triangles, one in each subtournament, with all edges between them oriented in the same direction. We disprove this conjecture by constructing, for every positive integer $d$, a tournament $T$ with a vertex partition $(X,Y)$ such that $χ(T[X]),χ(T[Y])\\ge d$, but no cyclic triangle in either part is complete to a cyclic triangle in the other. Our construction combines a graph of large girth and large chromatic number with a variant of Kneser graphs defined in terms…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-10-27","reactions":{}}