Navigating Quantum Missteps in Agent-Based Modeling: A Schelling Model Case Study

TL;DR

This work argues that directly translating agent-based models to quantum optimization frameworks (e.g., via QUBO) is structurally mismatched and unlikely to yield practical quantum advantage. Through Schelling's segregation case on lollipop networks, it shows that a count-first, structure-aware reformulation can dramatically outperform naive quantum reductions, establishing a concrete lower bound that quantum methods must beat. The authors demonstrate a fast classical algorithm that exploits network symmetries and show, both theoretically and empirically, that quantum speedups are unlikely under standard ABM access patterns. The study advocates reframing quantum ABM research around problem structure and rigorous classical baselines, outlining a pathway where quantum methods might truly outperform classical approaches when the problem architecture aligns with quantum primitives.

Abstract

Quantum computing promises transformative advances, but remains constrained by recurring misconceptions and methodological pitfalls. This paper demonstrates a fundamental incompatibility between traditional agent-based modeling (ABM) implementations and quantum optimization frameworks like Quadratic Unconstrained Binary Optimization (QUBO). Using Schelling's segregation model as a case study, we show that the standard practice of directly translating ABM state observations into QUBO formulations not only fails to deliver quantum advantage, but actively undermines computational efficiency. The fundamental issue is architectural. Traditional ABM implementations entail observing the state of the system at each iteration, systematically destroying the quantum superposition required for computational advantage. Through analysis of Schelling's segregation dynamics on lollipop networks, we demonstrate how abandoning the QUBO reduction paradigm and instead reconceptualizing the research question, from "simulate agent dynamics iteratively until convergence" to "compute minimum of agent moves required for global satisfaction", enables a faster classical solution. This structural reconceptualization yields an algorithm that exploits network symmetries obscured in traditional ABM simulations and QUBO formulations. It establishes a new lower bound which quantum approaches must outperform to achieve advantage. Our work emphasizes that progress in quantum agent-based modeling does not require forcing classical ABM implementations into quantum frameworks. Instead, it should focus on clarifying when quantum advantage is structurally possible, developing best-in-class classical baselines through problem analysis, and fundamentally reformulating research questions rather than preserving classical iterative state change observation paradigms.

Figures (7)

• Figure 1: (A) Initial state of Schelling's segregation model. (B) Final state of Schelling model with satisfied agents.
• Figure 2: An example of the welded tree problem. Two binary trees are welded together at their leaves through random connections, creating a complex middle region. Adapted from childs_quantum_forgetting_2023.
• Figure 3: $n$-dimensional hypercube graph, where vertices are binary strings and edges connect states differing in exactly one bit.
• Figure 4: Quantum walks effectively reduce the complex to a simple path with $2n+2$ vertices—one per level. The quantum superposition exploits symmetry to propagate efficiently without explicitly tracking individual paths. Adapted from childs2003exponential.
• Figure 5: An example of a $L_3^6$ lollipop network consisting of 4 red and 3 blue agents. White circles denote empty spaces in the network

Theorems & Definitions (7)

• Theorem 1: Output-size lower bound sipser1996introduction
• Theorem 2: Measurement budget and coherence klm
• Theorem 3
• proof
• Theorem 4
• proof
• Theorem 5