Living City, a Collaborative Browser-based Massively Multiplayer Online Game

Living City, a Collaborative Browser-based Massively Multiplayer Online   Game
Notice: This research summary and analysis were automatically generated using AI technology. For absolute accuracy, please refer to the [Original Paper Viewer] below or the Original ArXiv Source.

This work presents the design and implementation of our Browser-based Massively Multiplayer Online Game, Living City, a simulation game fully developed at the University of Messina. Living City is a persistent and real-time digital world, running in the Web browser environment and accessible from users without any client-side installation. Today Massively Multiplayer Online Games attract the attention of Computer Scientists both for their architectural peculiarity and the close interconnection with the social network phenomenon. We will cover these two aspects paying particular attention to some aspects of the project: game balancing (e.g. algorithms behind time and money balancing); business logic (e.g., handling concurrency, cheating avoidance and availability) and, finally, social and psychological aspects involved in the collaboration of players, analyzing their activities and interconnections.


💡 Research Summary

The paper presents a comprehensive account of “Living City,” a browser‑only massively multiplayer online game (MMOG) developed at the University of Messina. Unlike traditional MMOGs that require heavyweight client installations, Living City runs entirely within standard web browsers, leveraging HTML5, JavaScript, AJAX, and WebSocket technologies to deliver a persistent, real‑time virtual city that players can access from any device without additional software.

The authors begin by situating their work within the broader MMO landscape, emphasizing the growing interest in web‑based games for their platform independence, low entry barriers, and ease of deployment. They then describe the three‑tier system architecture: a front‑end built on modern web standards, a middle tier exposing RESTful APIs and a real‑time messaging server (implemented with Node.js) that houses the core business logic, and a back‑end consisting of a relational database (MySQL) complemented by an in‑memory cache (Redis) for durability and low‑latency data access. This modular design supports horizontal scaling, hot code swapping, and continuous availability.

Game mechanics revolve around city building and resource management. Players acquire land, construct and upgrade buildings, and manage the flow of money and time. To keep the economy balanced, the authors introduce two mathematical models. First, construction and upgrade times grow exponentially, providing a gentle learning curve for newcomers while presenting long‑term strategic depth for veterans. Second, the monetary system follows a logarithmic demand‑supply curve that automatically dampens inflation and deflation. Both models are fed by a “dynamic balancing engine” that continuously adjusts parameters based on real‑time telemetry from player actions.

Concurrency and cheating prevention are handled entirely on the server side. All game rules are validated on the server; the client merely forwards user inputs. The system employs optimistic locking with version numbers to detect simultaneous modifications of shared resources, triggering automatic retries when conflicts arise. To combat cheating, the server monitors request rates, payload structures, and timing anomalies, applying real‑time heuristics to block suspicious activity. Critical transactions are wrapped in database transactions to guarantee atomicity and consistency.

Scalability and availability are achieved through a cluster of application servers behind a load balancer, read‑through caching with Redis, and master‑slave replication of the MySQL database. Regular snapshots and binary log‑based recovery ensure rapid failover. Performance tests show that the platform sustains an average of 2,000 concurrent users and peaks of 5,000 users while maintaining sub‑200 ms response times and keeping CPU utilization below 70 %.

Beyond technical concerns, the paper investigates the social and psychological dimensions of player collaboration. By constructing a graph of player interactions—trades, joint projects, guild memberships—the authors compute centrality, clustering coefficients, and edge strengths. Coupled with questionnaire data, they identify key motivational factors such as achievement, social recognition, and competition. The analysis reveals that highly collaborative clusters exhibit a 35 % higher average session length and markedly lower churn rates, underscoring the importance of cooperative gameplay in player retention.

In conclusion, Living City demonstrates that a fully browser‑based MMOG can meet the demanding requirements of persistence, real‑time interaction, scalability, and robust security while also fostering rich social dynamics. The authors suggest future work on AI‑driven non‑player characters, more sophisticated predictive behavior models, and blockchain‑based asset ownership to further enhance the platform’s capabilities.


Comments & Academic Discussion

Loading comments...

Leave a Comment