TL;DR
This paper introduces the RA-Model, a novel taxonomy for classifying player activities in mobile MMORPGs based on two core motivations: Relationship and Achievement. By analyzing player data from a commercial mobile MMORPG, the authors demonstrate how this model can be used for player balance control and payment analysis, offering valuable insights for game design and maintenance.
What Problem Does This Paper Address?
The proliferation of mobile MMORPGs presents a challenge for game developers: understanding and categorizing the diverse player activities within these virtual worlds. While existing player taxonomies exist for PC MMORPGs (like Bartle's four types or Nick Yee's motivations), these may not directly translate to the unique characteristics of mobile gaming, such as fragmented play sessions. The authors aim to bridge this gap by proposing a simplified, yet effective, taxonomy specifically tailored for mobile MMORPGs, enabling better game design, balance, and monetization strategies.
Methods
The core of the paper is the RA-Model, which posits that player motivations in mobile MMORPGs can be distilled into two primary drivers: Relationship (desire to interact and socialize) and Achievement (desire to accomplish in-game goals).
- Data Collection and Preprocessing: Player data was extracted from the database of a commercial mobile MMORPG, "ZhanGuo." Player attributes were standardized to a 0-1 range to account for differing measurement units.
- Attribute Clustering: Cluster analysis was used to group player attributes into those reflecting Relationship (e.g., number of friends, family status) and those reflecting Achievement (e.g., weapon power, armor defense, skill level, win rate).
- Weight Calculation: Correlation analysis was employed to determine the weight of each attribute within its respective cluster (Relationship or Achievement).
- Score Calculation: Players' Relationship and Achievement scores were calculated as a weighted sum of their standardized attribute values.
- Player Taxonomy: Based on their Relationship and Achievement scores, players are classified into three types:
- Socializer: High Relationship, low Achievement.
- Moderator: Balanced Relationship and Achievement.
- Fighter: Low Relationship, high Achievement. The thresholds for these classifications (a and b) are game-specific.
Datasets
The study utilized data from a commercial mobile MMORPG called "ZhanGuo." This game has been running in China since 2005 and boasts over a million subscribers across 800 types of mobile phones. The authors specifically analyzed data from one week of activity on a single game server, involving 3712 active players. This dataset is presented as a new benchmark for analyzing player behavior in mobile MMORPGs, differing from questionnaire-based approaches used for PC games.
Key Insights and Ideas
- Simplicity is Key for Mobile: The RA-Model's strength lies in its parsimonious approach, reducing complex player motivations to just two core drivers, which is well-suited for the often fragmented nature of mobile gaming.
- Relationship Drives Payment: Counterintuitively, the study found that Socializers, despite having lower average payment amounts than Fighters in absolute terms, are significantly more likely to pay for the game. This suggests that strong social connections and community engagement are powerful drivers of player investment.
- Achievement for Survival, Relationship for Engagement: While Achievement is crucial for a player's survival and progression in MMORPGs, Relationship appears to be a more significant factor in fostering long-term attachment and willingness to spend.
- Practical Applications: The RA-Model is not just theoretical; it offers tangible benefits for game developers, enabling them to actively manage player balance and predict payment behavior, thereby optimizing game design and monetization.
Key Results
- The RA-Model successfully classified players into Socializer, Moderator, and Fighter types based on their Relationship and Achievement scores.
- In the "ZhanGuo" game, players generally had higher Achievement scores than Relationship scores, reflecting the game's design emphasis on combat and progression.
- Payment Analysis: Socializers, despite being 43% of the player base, contributed 15% of the total payment, with an average payment of 18.3. Fighters, representing 16% of players, contributed 56% of payments but had a much lower average payment of 1.8. This highlights the higher propensity of Socializers to pay, even if individual transactions are smaller.
- Balance Control: The model provides a framework for game companies to monitor and maintain the proportion of different player types, ensuring a stable and engaging game world.
Limitations and Future Work
- Game Specificity: The thresholds for classifying players into Socializer, Moderator, and Fighter (a=0.1, b=0.5) were determined based on the "ZhanGuo" game's experience. The authors acknowledge that these values will differ across various mobile MMORPGs.
- Generalizability: While the model is presented as generalizable to all mobile MMORPGs, further validation across a wider range of games is needed.
- Future Work: The authors plan to apply the RA-Model to more mobile MMORPGs to further verify its usability and practicality. They also intend to conduct more detailed analyses of player payment characteristics across different player types.
Summary updated