🎉 Celebrating 25 Years of GameDev.net! 🎉

Not many can claim 25 years on the Internet! Join us in celebrating this milestone. Learn more about our history, and thank you for being a part of our community!

Back to Artificial Intelligence

Looking for Related Work and feedback concerning "Approaches to Game AI under Consideration of Gameplay Mechanics"

Rockroot · 2019-07-12T10:20:49

Hello dear AI folk! I picked up a passion for Game AI a while ago and I wanted to join a community for a while now. This seems to be the right place to get cozy. Now, for my bachelors degree, I am supposed to write a research essay (about 20 pages). I chose to title it "Comparing Approaches to Game AI imder Consideration of Gameplay Mechanics". In the essay I am making bold statements, concerning the state of related work in the field. I want to know, if it holds up to reality. And what better way to find out, than asking the community of said reality? Quote IV. Related Work There are quite some sources to draw from, when entering the field of Game AI programming. “The AI Programmer’s Bookshelf“[1] is a good place to start a beginner‘s journey to this discipline of programmming. Next to the rather scientific analysis of the algorithms, there is alot of information to be collected from post mortems and anecdotal reports of AI hacks, tricks and work ethics to be found all over the internet, or comprised into books. On the other hand the field of game design is starting to create rules and tools for unified language. It is a requirement for research and communication of common patterns and conceptions in a way, that is re-usable as in science as in context of development. Lists of game design patterns and gameplay mechanics are yet incomplete and inhomogeneous in structure and detail. The information about any specific relation between types of gameplay and types of AI is vast, but barely exploited in form of scientific documentation. This lack of documentation is comprehensible for two reasons: First, AI is a selling point for alot of games. Non disclosure agreements keep developers closemouthed about the details of recent or current work. Also, the player community is carefully respected for finding exploits in a game’s AI and possibly breaking the game for them and others. Second, successful developers already do the right job and have no need for an abundant list of gameplay mechanics, game AI structures, or their correlation. Dedicated AI developers grow with experience at the hand of their productions, or by those of others in the form of post mortems. It is far more profitable for them to use their time applying the techniques, than documenting them. [1] http://alumni.media.mit.edu/~jorkin/aibooks.html Disclaimer: This is the first research paper I have ever done and it is work in progress. I feel that I suck at this. The main question is: Does the above statement hold up to reality? But please don't shy away from corrections or general advice. I would also like to share the completed work as soon as it is done, if anyone is interested. I am in dear need of some feedback, that I personally can grow by. Cheers

Artificial Intelligence Programming R&D

Started by Rockroot May 31, 2018 07:46 AM

11 comments, last by Jack_White 4 years, 11 months ago

Rockroot

Author

October 12, 2018 09:10 AM

Hello dear AI folk,

like I said in my Bachelor thesis post, I am done with my Bachelor thesis and the accomodating research project, that we discussed above. I wanted to thank you all for your previous feedback and to present you what came from it.

Spoiler

Originally the thesis was supposed to have 60 pages - my university decided to split it into 20 pages of research project and 40 pages of hands-on work (prototype) with documentation for all students...

I don't really understand why the separation, something about "People tend to forget the scientific workflow when they are not forced to do so within the research project ...

it felt like doing 2 BA thesises (thesees? thesis's?) at the same time, within a very constrained number pages ...

I hope it is an interesting read (20 pages), should you care to have a look. I was very satisfied with this one (more so than with the BA thesis), even though it did not receive any grade.

Again, feedback is very welcome, as I want to grow from it.

Anyway, I felt like this thread was incomplete, if I did not post the result in the end.

Thank you all for your suggestions! Even though they were frustrating at times, they really helped me to improve!

EDIT: The piece is written to give other students a shortcut to what architecture they could pick, so it is addressed at beginners!

PS. Both pieces seem rather Buzzfeed'esque to me. Maybe that's all you get in a Bachelor thesis?

732830_ThomasHaase_Research_GameAI.pdf

Jack_White

July 12, 2019 10:20 AM

8
Advanced issues found
▲

3

On 10/12/2018 at 12:10 PM, Rockroot said:

Hello dear AI folk,

like I said in my Geography essay questions, here I am done with my Bachelor thesis and the accomodating research project, that we discussed above. I wanted to thank you all for your previous feedback and to present you what came from it.

Reveal hidden contents

Originally the thesis was supposed to have 60 pages - my university decided to split it into 20 pages of research project and 40 pages of hands-on work (prototype) with documentation for all students...

I don't really understand why the separation, something about "People tend to forget the scientific workflow when they are not forced to do so within the research project ...

it felt like doing 2 BA thesises (thesees? thesis's?) at the same time, within a very constrained number pages ...

I hope it is an interesting read (20 pages), should you care to have a look. I was very satisfied with this one (more so than with the BA thesis), even though it did not receive any grade.

Again, feedback is very welcome, as I want to grow from it.

Anyway, I felt like this thread was incomplete, if I did not post the result in the end.

Thank you all for your suggestions! Even though they were frustrating at times, they really helped me to improve!

EDIT: The piece is written to give other students a shortcut to what architecture they could pick, so it is addressed at beginners!

PS. Both pieces seem rather Buzzfeed'esque to me. Maybe that's all you get in a Bachelor thesis?

732830_ThomasHaase_Research_GameAI.pdf

All these components of the game can be used to develop or analyze a complete game. While developing the game, many novice designers have no idea where to start. A finished game like World of Warcraft is so huge and vast that starting development of a holistic thing from scratch is an impossible task.