The Study of Procedural Content Generation in Video Gaming

The industry of video gaming is increasing becoming procedural content generation methods. The procedural algorithms that are used in the video gaming industry have presented a high degree of unpredictability. In that case, designers are expected to perform several tests and simulations until they get the right procedural system.

2. The work is motivating because it takes the advantage of using SBPCG in the analysis of terrains. For example, Stuerzlinge and Stachniak managed to use the stochastic local search algorithm, which helps with the finding of the acceptable deformation operations applied to the base terrain.

3.  The hypothesis presented by the authors is the development of an Automated Genetic Terrain Programming, which is an SBPCG technique for generating procedural terrains in video gaming. The method utilizes genetic programming as the search tool for automated procedural terrains. The technique is better than the previous methods, whereby aesthetic evaluation was performed by humans.

4. The authors have also managed to test the technique in the previous works, whereby the question regarding the creativeness of the technique was left open. The argument was that assessment of creativity should account for the performance of the software in terms of process and functionality.  Future improvements for the proposed technique entails the establishment of a comparison base for the GTPa creativity and diversity of the aesthetic terrain.

5.    The underlying problem that comes with the analysis of the presented technique is creativity.  Even though the analysis involving GTPa, technique have been presented in many publications, still creativity has been an issue. Small amounts of TPs have been generated with only a single metric for fitness function is included.

6.     The experimental techniques presented by the author include the evaluation of TPs, which means Test Parameters. The test was aimed at assessing the creativity and diversity of the terrain of the terrains that can be produced by the GTPa. This involved performance of series of tests, which entail the combination test parameters such as w_m, T_i, S_j,p_ak, and p_ei.

7.  Since the creativity analysis on the GTPa is still at the preliminary stage, the future directions of the research will be about the same issues.  Future studies will also involve the application of Ritchie’s criteria in implementing the creativity aspect of the design. Another further study will involve the use of classification system for aggregating terrains with their morphological similarity. In this way, it will be easy to assess phenotype diversity of the technique.

8.   Questions relating to the technique presented by the authors are as follows:

a. How many tests have been made with the GTPa whereby the fitness function involves the combination of the two metrics?

b. How is the analysis of the two metrics, edge length score and accessibility score is suitable to the design of the GTPa technique?

c. Is there any possibility of performing user study while classifying the creativity characteristics of the terrains?

9. The most important references cited in the paper were as follows:

a. M. J. Nelson and M. Mateas, “Towards automated game design,” in AI*IA ’07: Proceedings of the 10th Congress of the Italian Association for Artificial Intelligence on AI*IA 2007. Berlin, Heidelberg: Springer-Verlag, 2007, pp. 626–637.

b. C. Remo, “MIGS: Far Cry 2’s Guay On The Importance Of Procedural Content,” Website (accessed on Sep. 2011), 2008, http://www.gamasutra.com/PHP-bin/news\ index.php?story=21165.

c. J. Togelius, G. N. Yannakakis, K. O. Stanley, and C. Browne, “Search-based procedural content generation: A taxonomy and survey,” Computational

d. Intelligence and AI in Games, IEEE Transactions on, vol. 3,. 3, pp. 172 –186, sept. 2011.

e. S. Stachniak and W. Stuerzlinger, “An algorithm for automated fractal terrain deformation,” Computer Graphics and Artificial Intelligence, vol. 1, pp. 64–76, 2005.

f. T. J. Ong, R. Saunders, J. Keyser, and J. J. Leggett, “Terrain generation using genetic algorithms,” in GECCO ’05: Proceedings of 2005 conference on Genetic and evolutionary computation. NY, USA: ACM, 2005, pp. 1463–1470

10. The authors of the paper drew their ideas from the previous research.  They were motivated by the idea of the GTPa technique because the industry of video gaming is increasingly becoming a procedural content generation methods. Furthermore, the previous studies have been authorizing processes of trial and error.

11. The idea of the classification system in aggregating terrains will work best for the GTPa, depending on the level of the required creativity.

12.  The set of the data presented by the authors are thoroughly explained and reproducible. They are reproducible in the manner that the expected results are based on the previous findings.