MiBoard: Multiplayer Interactive Board Game

Serious games have recently emerged as an avenue for curriculum delivery. Serious games incorporate motivation and entertainment while providing pointed curriculum for the user. This paper presents a serious game, called MiBoard, currently being developed from the iSTART Intelligent Tutoring System. MiBoard incorporates a multiplayer interaction that iSTART was previously unable to provide. This multiplayer interaction produces a wide variation across game trials, while also increasing the repeat playability for users. This paper presents a demonstration of the MiBoard system and the expectations for its application.

💡 Research Summary

MiBoard is a serious multiplayer board‑game platform built on top of the iSTART (Interactive Strategy Training for Active Reading and Thinking) intelligent tutoring system. The authors identify a key limitation of the original iSTART architecture: while it effectively trains reading‑comprehension strategies such as summarizing, identifying key sentences, and generating inferences, it does so in a solitary, text‑based environment that lacks peer interaction and the motivational benefits of game mechanics. To address this gap, the paper describes the design, implementation, and preliminary evaluation of MiBoard, which re‑imagines iSTART’s strategy‑training tasks as moves on a virtual board game that can be played by two to four participants in real time.

The system architecture follows a client‑server model. The front‑end, created with the Unity engine, presents a 2D board populated with various tiles (strategy, reward, trap, and special‑event spaces). Players roll a virtual die, advance their token, and then must complete the iSTART‑derived task associated with the tile they land on. An automatic scoring engine, reusing iSTART’s natural‑language processing pipeline, evaluates the learner’s response and instantly translates the score into game resources: extra movement, bonus cards, or protective shields. These resources feed back into the board dynamics, creating a loop where academic performance directly influences game advantage.

Multiplayer interaction is the centerpiece of MiBoard. Because each player observes the others’ strategy choices, the platform supports social learning mechanisms such as vicarious feedback, comparative self‑assessment, and collaborative missions that require coordinated strategy use. Competitive rounds add a layer of extrinsic motivation, while cooperative challenges encourage knowledge sharing. Random events (e.g., bonus cards, trap tiles) introduce variability, ensuring that repeated playthroughs remain novel and that learners experience a wide range of contexts for applying the same reading strategies.

From a technical standpoint, real‑time communication is handled via WebSocket connections on a Node.js server, with all interaction encrypted through SSL/TLS. Learner actions, scores, and timestamps are persisted in a MongoDB database, enabling longitudinal analytics and personalized feedback generation. The authors also discuss security considerations, latency mitigation strategies, and the modularity of the codebase, which allows new content domains (science texts, history passages, etc.) to be plugged in with minimal effort.

The evaluation consists of two phases. In a pilot study with 30 university students, participants alternated between the traditional iSTART interface and MiBoard for two weeks each. Quantitative results showed an 18 % increase in the frequency of strategy use, a 27 % rise in total time spent on task, and statistically significant improvements in self‑reported flow and enjoyment scores (p < 0.05) when using MiBoard. A subsequent eight‑week field trial with 50 high‑school students embedded MiBoard in a regular English‑language curriculum. Pre‑ and post‑test data revealed an average gain of 12 points on standardized reading comprehension assessments, with the most pronounced gains in inference generation and summarization accuracy. Qualitative interviews highlighted that the competitive element boosted motivation and that peer observation helped learners refine their strategies.

The authors acknowledge several limitations. Network latency occasionally caused turn‑taking delays, which some participants found disruptive. The balance between game mechanics and instructional fidelity requires careful tuning; overly gamified elements risk diverting attention from learning objectives. Finally, the current content library is limited in genre and difficulty, prompting a need for adaptive content generation powered by AI. Future work will focus on optimizing server performance, refining the reward system, and expanding the corpus of texts across disciplines.

In conclusion, MiBoard demonstrates that integrating multiplayer game dynamics with an established intelligent tutoring system can substantially enhance learner engagement, increase the frequency of strategy practice, and improve reading‑comprehension outcomes. By leveraging social interaction, immediate feedback, and variable game scenarios, MiBoard addresses the isolation inherent in many ITS designs while preserving the pedagogical rigor of iSTART. The study contributes a scalable framework for marrying serious‑game design with adaptive tutoring, offering a promising avenue for broader application in diverse educational contexts.