Functional Customization in Looter Shooters and Roguelikes
Does deeper customization actually motivate players, or just give them more menus? I tested it across 10 looter shooters and roguelikes. Customization depth significantly drives competence and autonomy, but only in specific subcategories. Production implications below.
A games user research study deconstructing functional customization across 10 looter shooters and roguelikes, including Borderlands and Slay the Spire. The study applied Self-Determination Theory and Self-Efficacy Theory to test whether customization depth meaningfully affects player motivation, and what producers should take away.
- How does customization shape a player's motivational experience?
- In what ways do failure gameplay loops influence engagement with customization?
- Does customization meaningfully affect player confidence?
Looter shooters (no failure gameplay loops)
Roguelikes (failure gameplay loops)
Three-part customization taxonomy
Built and applied a framework dividing functional customization into offensive, abilities, and defensive subcategories. Used it to deconstruct each of the 10 games and rate each as high or low across the three SDT constructs.
180-participant survey
Qualtrics survey distributed via game-specific subreddits, Discord communities, SMU Guildhall, and SMU Game Club. 40 items per survey, operationalized to each game.
Validated scales
BPNSFS for SDT constructs (autonomy, competence, relatedness) and NGSES for self-efficacy. Cronbach's alpha used to confirm internal consistency before analysis.
Statistical analysis
One-way ANOVA, Welch's ANOVA where homogeneity of variance failed, and post-hoc subcategory analysis to isolate which customization types drove the significant relationships.
- Functional customization significantly affects competence and autonomy, but not relatedness or self-efficacy. The presence of failure loops alone didn't predict differences in motivation across genres.
- Offensive customization is the strongest driver of player competence. Games with more weapons, damage types, and combat styles had players reporting higher feelings of mastery.
- Abilities customization showed an inverted relationship. Games with fewer ability systems had higher competence scores, suggesting that too many skill trees or class options can work against perceived mastery rather than support it.
- Autonomy comes from breadth of customization, not any one category. No single subcategory was individually significant for autonomy, but the overall scope was.
- Unexpected outlier: Risk of Rain 2 had the highest autonomy mean despite its randomized item drops. Qualitative responses suggested players value real-time adaptation as much as pre-game planning.
- Prioritize offensive and ability customization depth if you're designing for player competence. Weapons, damage systems, combat styles, class specializations, and skill trees are the highest-signal levers.
- Watch for ability bloat. Adding more skill trees or class options past a threshold may undercut competence rather than enhance it.
- For autonomy, breadth across multiple customization categories matters more than depth in any single one. A game with moderate options across offensive, abilities, and defensive will likely feel freer than one with deep weapons customization but shallow everything else.
- Random elements don't necessarily reduce player agency. Risk of Rain 2 demonstrated that players can derive a strong sense of control from adaptation rather than planning.
- 180 total participants spread across 10 titles; per-game samples didn't reach the calculated 385 target for 95% confidence with 5% margin of error.
- Recruitment via gaming subreddits, Discord communities, and Guildhall channels introduces self-selection bias toward core/dedicated players.
- The three-part taxonomy was useful for structuring deconstruction but didn't reliably predict player constructs. The paper does not recommend treating it as predictive.
Full thesis paper and defense slides below.