RUS  ENG
Full version
JOURNALS // Doklady Rossijskoj Akademii Nauk. Mathematika, Informatika, Processy Upravlenia // Archive

Dokl. RAN. Math. Inf. Proc. Upr., 2025 Volume 527, Pages 206–216 (Mi danma679)

SPECIAL ISSUE: ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING TECHNOLOGIES

NP-completeness of Hanabi game with minimal parameters

A. A. Onoprienko

National Research University Higher School of Economics, Moscow

Abstract: We study the algorithmic complexity of the cooperative card game Hanabi. The feature of Hanabi is that players see each other's cards but not their own, and exchange information through hints. Even in the model with one player who has full information about the deck, Hanabi remains NP-hard. We found the minimal parameters of the game that preserve NP-hardness. If these parameters are further reduced, the game turns out to be solvable in polynomial time.

Keywords: algorithmic game theory, computational complexity, NP-completeness, Hanabi.

UDC: 004.8

Received: 12.08.2025
Accepted: 22.09.2025

DOI: 10.7868/S2686954325070173



Bibliographic databases:


© Steklov Math. Inst. of RAS, 2026