Complexity Thoughts: Issue #78
Unraveling complexity: building knowledge, one paper at a time
If you find value in #ComplexityThoughts, consider helping it grow by subscribing and sharing it with friends, colleagues or on social media. Your support makes a real difference.
→ Don’t miss the podcast version of this post: click on “Spotify/Apple Podcast” above!
Human behavior
The paradox of intervention: Resilience in adaptive multirole coordination networks
Criminal networks are often targeted with focused interventions to weaken their structure and disrupt operations. However, using empirical data from the Dutch National Police, we show that such interventions can have the opposite effect, making criminal networks more resilient and better coordinated. Our study combines real-world data with computational models that capture how individuals adapt their roles in response to external pressure. The findings challenge common assumptions about how complex systems respond to disruption and have implications far beyond crime, affecting how we think about resilience in ecosystems, organizations, and technological networks. By revealing the unintended consequences of well-meaning interventions, our work urges a more nuanced approach to disrupting complex adaptive systems.
Read also this Commentary by Matjaz Perc discussing the unintended consequences of well-intended interventions.
“these findings illustrate how traditional law enforcement strategies, by focusing on simple disruption, may inadvertently foster more resilient and dangerous criminal organizations”
Complex adaptive networks exhibit remarkable resilience, driven by the dynamic interplay of structure (interactions) and function (state). While static-network analyses offer valuable insights, understanding how structure and function coevolve under external interventions is critical for explaining system-level adaptation. Using a unique dataset of clandestine criminal networks, we combine empirical observations with computational modeling to test the impact of various interventions on network adaptation. Our analysis examines how networks with specialized roles adapt and form emergent structures to optimize cost–benefit trade-offs. We find that emergent sparsely connected networks exhibit greater resilience, revealing a security–efficiency trade-off. Notably, interventions can trigger a “criminal opacity amplification” effect, where criminal activity increases despite reduced network visibility. While node isolation fragments networks, it strengthens remaining active ties. In contrast, increasing a node’s connectivity (analogous to social reintegration) can unintentionally boost criminal coordination, increasing activity or connectivity. Failed interventions often lead to temporary functional surges before reverting to baseline. Surprisingly, stimulating connectivity destabilizes networks. Effective interventions require precise calibration to node roles, connection types, and external conditions. These findings challenge conventional assumptions about connectivity and intervention efficacy in complex adaptive systems across diverse domains.
Emergence of simple and complex contagion dynamics from weighted belief networks
Social contagion is a ubiquitous and fundamental process that drives individual and social changes. Although social contagion arises as a result of cognitive processes and biases, the integration of cognitive mechanisms with the theory of social contagion remains an open challenge. In particular, studies on social phenomena usually assume contagion dynamics to be either simple or complex, rather than allowing it to emerge from cognitive mechanisms, despite empirical evidence indicating that a social system can exhibit a spectrum of contagion dynamics—from simple to complex—simultaneously. Here, we propose a model of interacting beliefs, from which both simple and complex contagion dynamics can organically arise. Our model also elucidates how a fundamental mechanism of complex contagion—resistance—can come about from cognitive mechanisms.
Slower searching yields higher efficiency: A case study of taxi drivers
How do humans optimize search strategies in complex, real-world environments? By analyzing Global Positioning System (GPS) data from taxi drivers in three large cities, this study uncovers a counterintuitive finding: More efficient drivers search at slower speed and make more short-distance turns, improving their success. Notably, search efficiency appears to be an individual trait, with efficient drivers consistently outperforming others over time. However, only about 10% of drivers adopt this optimal strategy, earning nearly 20% more than the average driver. These findings provide insights into human decision-making, demonstrating that deliberate, slower search strategies can outperform faster ones—even in fast-paced, competitive settings. This research advances our understanding of cognitive and behavioral strategies, with implications for psychology, behavioral economics, and operations research.
The movement patterns of animals while searching for food have been studied extensively in recent decades. However, although human search behavior has existed since the beginning of civilization, not much is known about human search patterns, particularly regarding strategies that yield higher efficiency. Although most humans no longer need to gather and hunt in the wild, human searching remains prevalent in modern times. A common example of human searching is performed by taxi drivers looking for passengers in a city. Here, we analyze GPS data of taxi drivers in three major cities over different time periods and find that when drivers search for passengers, the higher is their efficiency, the slower is their searching, and they tend to make more short-distance turns during the search. Our study further indicates that individuals are characterized by a specific level of efficiency, and thus, efficient drivers are consistently efficient across different days as they follow their own search strategies. Interestingly, only about 10% of drivers adopt the most efficient strategy, earning nearly 20% more than the average driver. Our findings shed light on human search behavior, a fundamental aspect of human decision-making in competitive and fast-paced environments.
Early insight into social network structure predicts climbing the social ladder
While occupying an influential position within one’s social network brings many advantages, it is unknown how certain individuals rise in social prominence. Leveraging a longitudinal dataset that tracks an entirely new network of college freshmen (N = 187), we test whether “climbing the social ladder” depends on knowing how other people are connected to each other. Those who ultimately come to occupy the most influential positions exhibit early and accurate representations of their network’s general, abstract structure (i.e., who belongs to which communities and cliques). In contrast, detailed, granular representations of specific friendships do not translate into gains in social influence over time. Only once the network stabilizes do the most influential individuals exhibit the most accurate representations of specific friendships. These findings reveal that those who climb the social ladder first detect their emerging network’s general structure and then fine-tune their knowledge about individual relationships between their peers as network dynamics settle.
Epidemiology models explain rumour spreading during France’s Great Fear of 1789
The Great Fear of 1789, a wave of panic and unrest in rural France fuelled by the spreading of rumours, was an important moment at the onset of the French Revolution, marking the collapse of feudalism and the rise of the new regime1. The Great Fear provides a vivid example of the role the spreading of rumours has in driving political changes that might be relevant today2,3. Here, we collect existing historical records related to the Great Fear and use epidemiology tools and models4 to reconstruct the network of its transmission from town to town. In this way, we quantify the spatiotemporal spread of the rumours and compute key epidemiological parameters, such as the basic reproduction number. Exploiting information on the structure of the road network in eighteenth century France5, we estimate the most probable diffusion paths of the Great Fear and quantify the distribution of spreading velocities. By endowing the nodes in our reconstructed network with indicators related to the institutional, demographic and socio-economic conditions of the time6, including literacy, population size, political participation, wheat prices7,8, income and ownership laws9, and the unequal distribution of land ownership, we compute factors associated with spread of the Great Fear. Our analysis sheds light on unresolved historiographic issues on the significance of the Great Fear for the French Revolution, providing a quantitative answer to the unresolved debate between the role of emotions and rationality in explaining its diffusion.
Decoupling geographical constraints from human mobility
Driven by access to large volumes of movement data, the study of human mobility has grown rapidly over the past few decades. The field has shown that human mobility is scale-free, proposed models to generate scale-free moving distance distributions and explained how the scale-free distribution arises. It has not, however, explicitly addressed how mobility is structured by geographical constraints, such as how mobility relates to the outlines of landmasses, lakes and rivers and the placement of buildings, roadways and cities. On the basis of millions of moves, we show how separating the effect of geography from mobility choices reveals a power law spanning five orders of magnitude. To do so, we incorporate geography via the pair distribution function, which encapsulates the structure of locations on which mobility occurs. By showing how the spatial distribution of human settlements shapes human mobility, our approach bridges the gap between distance- and opportunity-based models of human mobility.
Online misinformation promotes distrust in science, undermines public health, and may drive civil unrest. During the coronavirus disease 2019 pandemic, Facebook—the world’s largest social media company—began to remove vaccine misinformation as a matter of policy. We evaluated the efficacy of these policies using a comparative interrupted time-series design. We found that Facebook removed some antivaccine content, but we did not observe decreases in overall engagement with antivaccine content. Provaccine content was also removed, and antivaccine content became more misinformative, more politically polarized, and more likely to be seen in users’ newsfeeds. We explain these findings as a consequence of Facebook’s system architecture, which provides substantial flexibility to motivated users who wish to disseminate misinformation through multiple channels. Facebook’s architecture may therefore afford antivaccine content producers several means to circumvent the intent of misinformation removal policies.
Artificial Intelligence
Discovering network dynamics with neural symbolic regression
Network dynamics are fundamental to analyzing the properties of high-dimensional complex systems and understanding their behavior. Despite the accumulation of observational data across many domains, mathematical models exist in only a few areas with clear underlying principles. Here we show that a neural symbolic regression approach can bridge this gap by automatically deriving formulas from data. Our method reduces searches on high-dimensional networks to equivalent one-dimensional systems and uses pretrained neural networks to guide accurate formula discovery. Applied to ten benchmark systems, it recovers the correct forms and parameters of underlying dynamics. In two empirical natural systems, it corrects existing models of gene regulation and microbial communities, reducing prediction error by 59.98% and 55.94%, respectively. In epidemic transmission across human mobility networks of various scales, it discovers dynamics that exhibit the same power-law distribution of node correlations across scales and reveal country-level differences in intervention effects. These results demonstrate that machine-driven discovery of network dynamics can enhance understandings of complex systems and advance the development of complexity science.
Results for ecological and epidemic dynamics are very interesting and show how this method can be considered a valid competitor for other state-of-the-art method.
I should write a dedicated post about this in the future!
→ Please, remind that if you find value in #ComplexityThoughts, you might consider helping it grow by subscribing, or by sharing it with friends, colleagues or on social media. See also this post to learn more about this space.