Researcher Corrects Categorical Framework for Modeling Strategic Conflict

A substantially updated mathematical paper reveals crucial errors in an earlier attempt to map game theory's conflict resolution models onto category theory—and offers a corrected framework that reveals why preferences behave differently from other structural elements.

The Core Problem#

The original framework tried to treat both the mechanics of conflict (moves and states) and players' preferences as categorical functors—mathematical objects that preserve structure. The revision shows this was fundamentally mistaken. While decision-maker labels (who controls each move) can indeed be represented functorially as mappings into a free monoid, preferences cannot work the same way.

What Works Categorically#

The corrected approach models states and moves as a free category built on a quiver (directed graph). Moves are labeled by the player controlling them, creating paths through the conflict space. Crucially, legal sequences used to determine what alliances can reach—paths where no player makes two consecutive moves—can now be formally captured as paths whose label patterns contain no immediate repetition.

Where Preferences Break Down#

Mapping a player's preferences into ordered sets requires those preferences to be transitive (consistent). Even when that holds, preferences extend to become a proper functor only under an extremely restrictive condition: every single move by any player must be weakly beneficial for the player being analyzed. When this rare situation occurs, three different stability concepts—symmetric metarationality, sequential stability, and general metarationality—all collapse into the single concept of Nash stability.

Instead, preferences must be treated as a selection of moves representing improvements: a subquiver rather than a functor.

Practical Implications#

The framework is tested on the Prisoner's Dilemma and examined across different conflict models. The analysis finds that Nash stability is preserved under model embeddings (it "reflects" mathematically), while the other stability concepts are neither preserved nor reflected—meaning they behave less predictably when comparing different game structures formally.

The revision systematically catalogs every correction from the original version, offering a foundation for future categorical approaches to strategic interaction.