Graph-Based Cops and Robbers Games

Graph-based Cops and Robbers Games constitute a class of pursuit–evasion problems played on abstract graphs, in which a set of pursuers (cops) attempts to capture an evader (robber) moving along vertices and edges. Since its introduction in the early 1980s, this framework has shaped discrete mathematics and theoretical computer science while informing applied fields such as network security, robot motion planning and epidemiology. The classical variant assumes perfect information and alternating moves on a fixed, static graph, with fundamental parameters including the cop number—the minimum number of cops required to guarantee capture—and the capture time—the fewest rounds needed under optimal play. Extensions encompass invisible or partially observable robbers, stochastic or “drunken” movements, directed and dynamic graph models, and analyses on special classes such as planar, 1-planar and retract graphs. These studies yield structural bounds for sparse and dense networks, complexity classifications of decision problems, and quantitative measures of visibility and randomness in pursuit strategies. Emerging work further bridges to probabilistic combinatorics, control theory and real-time adaptive systems, underscoring both the theoretical depth and practical impact of these games.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

Recent studies on dynamic graphs have extended the Cops and Robbers paradigm to time-evolving networks. In the offline case, full knowledge of the evolution sequence enables polynomial-time algorithms to determine k-cop-win conditions, while in the online scenario adaptive pursuit protocols address uncertainty and establish near-tight bounds on the required number of cops relative to graph sparsity. Another prominent line of inquiry examines “tipsy” variants in which cop and robber interleave directed and random moves. Analyses on infinite grid graphs and infinite trees quantify how biased random walks affect capture probabilities and expected game lengths, revealing that stochastic perturbations can frustrate pursuit yet still admit efficient bounding strategies. A third direction explores pursuit–evasion on oriented graphs via strong and weak move models: despite preserving some classical invariants, these directed variants exhibit novel complexity behaviours and demand refined structural characterisations—through retracts and subdivisions—to determine cop numbers under orientation constraints.

Graph-Based Cops and Robbers Games publication trend

The graph below shows the total number of articles in graph-based cops and robbers games across all publications each year (not limited to Nature Index journals).

Technical terms

Cop number: The minimum number of cops needed to guarantee capture of the robber on a given graph under optimal play.

Capture time: The number of rounds required for the cops to catch the robber when both sides play optimally.

Pursuit–evasion: A class of games and problems modelling the strategic interaction between one or more pursuers and one evader on a given domain.

Dynamic graph: A graph whose vertex set or edge set changes over discrete time steps, modelling time-dependent networks.

References

1. The Role of Visibility in Pursuit/Evasion Games. Robotics (2014).
2. Chasing a Drunk Robber in Many Classes of Graphs. Dynamic Games and Applications (2022).
3. Cops and Robbers on Dynamic Graphs: Offline and Online Case. Discrete Mathematics & Theoretical Computer Science (2023).
4. Tipsy cop and tipsy robber: Collisions of biased random walks on graphs. Theoretical Computer Science (2024).
5. Cops and robber on variants of retracts and subdivisions of oriented graphs (Brief Announcement). Procedia Computer Science (2023).
6. On 1-planar graphs with bounded cop-number. Theoretical Computer Science (2025).