Influencing Player Tension Using Spatial Composition: Revisiting and Extending the Yuan – Wang Framework
Jiafan LU
Level Design
Influencing Player Tension Using Spatial Composition: Revisiting and Extending the Yuan – Wang Framework
I aimed to quantify how spatial composition in video game levels influences player tension. While tension is traditionally shaped by a designer’ s intuition, I wanted to explore whether it can be systematically predicted before players even step into a space. of distinct spatial nodes varying in density, openness, and scale. I collected player responses through biometric data( iMotions facial expression analysis) and post-playthrough surveys, then compared the predicted tension curves against actual player experience.
My study builds on two prior SMU Guildhall theses by Canny Yuan and Yidi“ Edith” Wang. I integrated Yuan’ s Spatial Tension formula with Wang’ s Blocked Score method, replacing the subjective Density component with a grid-based objective calculation( Blocked Score). The resulting formula, Spatial Tension Score( BS), allows designers to predict tension using measurable, reproducible inputs rather than personal judgment.
The results confirmed that the integrated formula improves predictive accuracy at most spatial nodes while also revealing its boundary conditions in non-conventional spaces such as narrow passageways and partially explored rooms. This project took approximately 11 months to complete and offers level designers a practical tool for objectively evaluating spatial tension during production.
To validate this approach, I built a standalone Starfield level titled“ Remnant of Tiangong,” structured as a sequence
48 LEVEL DESIGN