Beyond Asking: A Pipeline for Personalized Game Generation that Reads Players from Behavior

2026-08-17

AI Summary

This study addresses the challenges of player trait verification and behavioral ambiguity in personalized game generation by proposing an opportunity-aware decision representation that decouples preferences from opportunities. We construct a verifiable synthetic player benchmark based on explicit parameters and employ few-shot large language model reasoning to infer traits from behavior, thereby driving a closed-loop adaptive difficulty evaluation. Experimental results demonstrate that this approach significantly outperforms baselines in trait inference and achieves effective closed-loop adaptation. Furthermore, preliminary evidence confirms the feasibility of transferring this framework to real players. Collectively, this work establishes a reliable verification framework and adaptive mechanism for personalized game generation, offering a robust solution to validate player models and optimize gameplay experiences dynamically.

Abstract

Personalized game generation requires inferring a player's abilities and behavioral style from how they play. Large language models have made this inference more attainable than ever: an LLM can read a raw gameplay transcript and produce a fluent, plausible profile of the player. Plausible, however, is not verified, and verification is precisely what the field lacks: latent traits are unobservable; questionnaires provide noisy proxies and become circular when self-reports are used to validate behavior-based inference; and behavior itself is ambiguous without context -- a player who never collects an item may not want it, or may never have had the chance. We address both problems. First, we construct a synthetic player population whose traits are ground truth by construction: each trait is an explicit bot parameter, accepted only after controlled manipulation produces consistent, trait-specific behavioral change. Unlike prior parameter-recovery work that inverts a known decision model, our benchmark evaluates policy-agnostic inference from behavioral transcripts alone. Second, we introduce an opportunity-aware decision-moment representation that disentangles preference from the chance to express it; ablating it selectively degrades opportunity-dependent traits. On this benchmark, few-shot LLM inference outperforms embedding- and rule-based baselines on most traits, though feature-based supervised regressors remain stronger overall. Finally, we close the loop: inferred profiles drive difficulty adaptation, evaluated against ground-truth references and mismatched-profile controls, and an exploratory human study examines whether these findings transfer to real players.

Problem

• Personalized Game Generation
• Player Profiling
• Behavioral Inference
• Latent Traits Verification
• Opportunity Ambiguity

Innovation

• Synthetic Player Benchmark
• Opportunity-aware Representation
• Behavioral Inference
• Personalized Game Generation
• Closed-loop Validation