Game theory proves good engineering makes good neighbours

Researchers in the School of Computing and Augmented Intelligence at Arizona State University, part of the Ira A Fulton Schools of Engineering, are using game theory to suggest ways to improve the design of shared infrastructure systems.

Picture a sun-scorched neighbourhood in the middle of summer. A small canal cuts between homes, quietly delivering water to yards via an old method: flood irrigation. It’s cheap, it’s effective and it absolutely depends on neighbours working together. But what happens when that cooperation breaks down?

Associate professor of industrial engineering Paul Grogan’s (pictured above, at right) latest research uses game theory and computer modelling to better understand how people make decisions in systems that require collaboration to function. His recent project focuses on one specific example: flood irrigation in Arizona neighbourhoods, where water from a shared canal must be jointly managed and maintained by nearby households.

If even one household drops out, for example by switching to a private drip-irrigation system, it increases costs and workload for everyone else. Too many defections and the system collapses. ‘This setup is a perfect example of collaborative infrastructure where the success of the system depends not just on good engineering but on sustained human cooperation,’ Grogan said.

To untangle the web of neighbourly incentives, Grogan and his team used a classic scenario from game theory known as the stag hunt. In this model, players choose between a high-reward option that requires teamwork, called hunting a stag, and a low-risk, lower-reward option they can do alone, referred to as hunting a hare. In this philosophical problem, both hunters know they could catch the stag if they cooperate, while working alone they can each easily obtain the far less valuable hare.

In the irrigation context, ‘hunting the stag’ means sticking with the shared canal system. The benefits include low costs and community goodwill, provided everyone participates. ‘Hunting the hare’ is switching to a private system, a choice that’s safe, but more expensive and isolating.

The research team tested different ways of laying out the irrigation infrastructure to see which encouraged the most cooperation. They used simulations and economic modelling to explore how trust and influence flow between neighbours depending on how the system is physically designed.

Their key finding? The way infrastructure is laid out plays a huge role in whether people choose to collaborate or go it alone.

Two network designs stand out. The bus topology, with houses connected to the canal like stops on a bus line, offers the highest economic efficiency. It keeps costs low by sharing infrastructure efficiently. But it isn’t perfect. Downstream users still depend on upstream neighbours, making trust fragile.

The star topology, on the other hand, places every household on equal footing. Everyone connects directly to the water source, reducing dependency and levelling influence. While slightly more expensive, this layout proves the most stable. Cooperation is more resilient and neighbours are more likely to work together.

Grogan’s research shows that better cooperation isn’t just a matter of better rules or stronger enforcement. Sometimes, it’s about smarter design.

The research also highlights student leadership. The study was co-authored by Mobin Zarreh (pictured above, at left), an industrial engineering doctoral student in the Collective Design Lab that Grogan leads at ASU. Zarreh developed the simulations that validated the theoretical models and gave the study its predictive power.

‘Mobin’s contributions went far beyond implementation,’ Grogan said. ‘He shaped the modelling framework and brought new ideas to the table at every stage.’

Grogan notes that Zarreh introduced a high level of focus on water resources and their management to the project. He met with representatives from utility company Salt River Project and attended the 2025 Arizona Groundwater Policy Conference. This involvement is not unusual in Grogan’s lab, where students take the lead in exploring real-world applications of complex systems design.

‘It was exciting to see how a theoretical model could be directly applied to something so tangible in our own communities,” Zarreh said. ‘This work deepened my appreciation for how engineering design can influence not just systems but also the way people interact and cooperate.’

Grogan’s flood irrigation study might seem like a hyperlocal case – just a few neighbours in Phoenix sharing a water system. But its implications are global.

As resources such as water become increasingly strained, infrastructure systems everywhere will depend more on cooperation between users. Grogan’s work shows that if we want people to work together, we shouldn’t just ask them nicely. We should build systems that make collaboration the easiest and most rewarding choice.

The research has been published in Water Resources Management.