Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


Dissertation.pdf (4.21 MB)

ETD Abstract Container

Abstract Header

Procedural Content Generation for Computer Games

Shi, Yinxuan
http://rave.ohiolink.edu/etdc/view?acc_num=osu1469147321

Abstract Details

2016, Doctor of Philosophy, Ohio State University, Computer Science and Engineering.
Procedural Content Generation (PCG) is no new concept for the gaming industry. From early games like Rogue (1980) and The Sentinel (1986) to more recent games like Diablo III (2012) and Path of Exile (2013), PCG is heavily used in dungeons, quests, mini bosses and even storyline creation. The advantages PCG offers is not just limited to empowering game designers with fast content prototype/creation, but can also provide in-game adaptation to player’s response and small memory footprint. While there is much research on PCG, few results contribute to the evaluation: Does the generated content makes the game more interesting/fun? To answer this question, we examine two applications of PCG. One is level creation and another is visual content creation such as crowds. For level creation, the existing techniques mainly focus on map/terrain generation. In games where the player either avoids or engages in combat against hostile targets, the player’s experience involves other aspects such as enemy and resource placement and navigation. The problem of creating a fun level can be formulated into searching for a good combination of these aspects. This leads to two problems: 1. How to evaluate the fun of a level? 2. How to constrain/sample the parameter space to produce a viable result in limited time? We tackle the first problem by placing a pseudo player into the level. A damage function is proposed to encode the flux of damage at every point in space throughout the level. For a shooter game, we work under the premise that there exists a path that is optimal in some sense through this damage field (i.e., there exists a path that would inflict the least amount of damage on the player). For a strategy game, we assume there is an optimal strategy for choosing paths for a small team to cross the damage field. With three different metrics which we defined, we are able to analyze a level by analyzing the optimal path(s). However, this search is NP. For the second problem, consider a level with a given terrain and entry and exit positions. All the possible configurations for enemies and resource placement are infinite. To better sample the parameter space, we lay down n candidate locations for enemies/resources. The problem is then transformed into a combinatorial problem. We divide the level by a grid and solve each grid cell for a fun enemy and cover combination. Rather than finding the optimal configuration out of 2^n possibilities, we treat each grid as a tile, with a precomputed tile set, we are able to obtain a fun level by finding a fun `tiling’ representation. The second application for PCG is the visual content. Visual realism and plausibility are the top criteria for assessing immersive experience in games. Here we investigate the representative distribution of body shapes when simulating crowds in games. Achieving representative and visually plausible body-shape variation while optimizing available resources is an important goal. We present a data-driven approach to generating and selecting models with varied body shapes, based on body measurement and demographic data from the CAESAR anthropometric database. With a perceptual study to explore the relationship between body shape, distinctiveness for bodies close to the median height and girth, we found that the most salient body differences are in size and upper-lower body ratios, in particular with respect to shoulders, waists and hips. Based on these results, we propose strategies for body shape selection and distribution that we have validated with a lab-based perceptual study. Finally, we demonstrate our results in a data-driven crowd system with perceptually plausible and varied body shape distribution that can be used in games.
Roger Crawfis (Advisor)
Yusu Wang (Committee Member)
Eric Fosler-Lussier (Committee Member)
Neelam Soundarajan (Committee Member)
148 p.
Computer Science
Procedural Content Generation, Computer Games

Recommended Citations

Citations

  • Shi, Y. (2016). Procedural Content Generation for Computer Games [Doctoral dissertation, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1469147321

    APA Style (7th edition)

  • Shi, Yinxuan. Procedural Content Generation for Computer Games. 2016. Ohio State University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1469147321.

    MLA Style (8th edition)

  • Shi, Yinxuan. "Procedural Content Generation for Computer Games." Doctoral dissertation, Ohio State University, 2016. http://rave.ohiolink.edu/etdc/view?acc_num=osu1469147321

    Chicago Manual of Style (17th edition)

osu1469147321
1,006
© 2016, some rights reserved.Creative Commons License Procedural Content Generation for Computer Games by Yinxuan Shi is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License. Based on a work at etd.ohiolink.edu.
This open access ETD is published by The Ohio State University and OhioLINK.