Tag: Rogue-like

Book Review: Procedural Generation in Game Design

From the first line of code I wrote, I have always been fascinated by procedural content generation (PCG) and the near-limitless potential it seems to hold for game development. Like so many other new developers, my first project was (of course) a wildly ambitious rogue-like. Needless to say, it didn’t quite pan out.

However, despite its challenges and limitations, I did keep my fascination for PCG and consider it a wonderful tool when applied correctly. Recently, I have been trying to read up on PCG while working on SKALD and a major gripe for me has been the lack of good literature regarding the subject. No wonder then, that I was very pleased to pick up a copy of Procedural Generation in Game Design.

Cover for the book Procedural Generation in Game Design

Procedural Generation in Game Design is a book consisting of 27 chapters (who read like essays) from different industry professionals. The book is edited by Tanya X. Short (creative director of Kitfox Games ) and Tarn Adams (co-creator of Dwarf Fortress) with a preface by Derek Yu (creator of Spelunky) .

On its back cover, the book lists the following four features:

  • Introduces the differences between static/traditional game design and procedural game design
  • Demonstrates how to solve or avoid common problems with procedural game design in a variety of concrete ways
  • Includes industry leaders’ experiences and lessons from award-winning games
  • World’s finest guide for how to begin thinking about procedural design

The book is divided into four sections: “Procedural Generation”, “Procedural Content”, “Procedural Narrative” and “The Procedural Future”.  Each section contains a collection of chapters that, more or less, share a common thread.

The topics covered in the different chapters is quite varied and include (among others): “When and Why to Use Procedural Generation”,  several procedural level design case studies, “Ethical Procedural Generation”, puzzle design, “Audio and Composition”, “Story and Plot Generation” and “Algorithms and Approaches”.

Though varying in both length and depth, all the individual chapters are quite good  with some even being excellent.

Mentioning a few, Mark R. Johnson’s (creator of Ultima Ratio Regnum) short chapter on meaning in PCG is great. So is Brian Bucklew’s (Freehold Games) chapter on “Algorithms and Approaches” (though I wish it had been longer). The entire section on procedural narrative is excellent as well, with the chapters by Ben Kybartas (Delft University of Technology) and Emily Short (check out her excellent blog) standing out in particular.

My major criticism of this book, however, is that it struggles in creating a coherent presentation and progression of content.

In several cases there seems to be a mismatch between chapter- and section topics. For instance: “Algorithms and Approaches” is oddly placed in the section called “Procedural Futures” even though the subject of the chapter is to give an overview of classical techniques used in PCG. In fact, I find that only the section called “Procedural Narrative” manages to maintain a strong coherency between the topics of its chapters.

As I have mentioned, I also find that the chapters vary somewhat in how deeply they explore their chosen topics. While a few read like abbreviated academic papers, others (the majority) feel more like blog posts. This is not to say that the quality of the content is poor: The chapters are written by highly talented game developers and provide inspiring insights into several well-known PCG-heavy indie game titles.

However, each chapter appears to have been written in isolation with only minimal direction concerning content. I find that the progression of content and relative amount of space given to each subject is also such that the book seems a bit underwhelming despite its 300+ pages. This is perhaps somewhat compounded by each author spending a few paragraphs talking about themselves and their project. Not that this is wrong, but it does dilute the PCG-specific content of the book.

The result is that the book does not live up to its full potential and promise of providing  the “World’s finest guide for how to begin thinking about procedural design”. Furthermore I find that this makes it hard to see who the books intended audience is: New developers may find the coverage of subjects incomplete while experienced developers may find it somewhat superficial.

Though the execution is far from perfect, the book (arguably) does mostly deliver on its promised features. And for all my criticisms, I did enjoy reading the individual chapters (or essays if you will). Therefore, I do recommend picking up this book if you’re looking for an interesting collection of individual essays concerning PCG by leading indie game developers. However, if you are looking for a comprehensive guide and introduction to PCG I doubt that this book alone will suffice.

You can pick up Procedural Generation in Game Design at Amazon for $49.77 (320 pages paperback, with grey-scale illustrations).

I was tipped off about this book by the very talented Filip Hráček.

Did you find this book review helpful? If you have any questions or comments, please get in touch. Also, follow Scape-IT and SKALD on Twitter for all things RPG and geeky!

 

 

 

Design Goals for the SKALD Engine

The SKALD Roleplaying System consists of three components:

  • A set of RPG rules usable for pen-and-paper as well computer RPGs.
  • A game engine for making gamebooks, interactive fiction and text-heavy roleplaying games.
  • The games published using the SKALD engine.

The current focus of the project is to finish the game engine. This post will attempt to summarize the design goals for the engine. For more information about the project in general, see the welcome post.

NOTE: I’m writing this post as much for myself as anyone else. As such, this post will be highly prone to edits as i update and change the design goals.

SKALD is a Storytelling Tool

First and foremost, I want to create something that will allow me to tell stories. I was a pen-and-paper RPG game master for years but as I grew older, holding together a gaming group became near impossible. I began to miss the creative outlet that GMing represented. This was been a big driving force in deciding to develop a game engine for text based roleplaying games and gamebooks.

Specifically, I chose text-heavy games because I feel comfortable with using text as a narrative device. Text is easy to add (if you don’t consider the hardships of writing prose) and very flexible. Flexibility was another important point as it gives me the option of choosing between an infinite variety of settings: From sci-fi and fantasy to historical and educational. This was an important rational for beginning this project by creating a setting-agnostic game engine.

SKALD is a roleplaying game

In addition to creating a system for writing choose-your-own-adventure style gamebooks I also want SKALD to be rooted in a proper RPG system. At a minimum the SKALD game engine must allow for the following game mechanics:

  • a character creation and advancement system complete with skills and feats
  • skill checks
  • tactical combat (in one way or another)
  • looting, buying and selling items
  • magic and / or psionics

An important part of implementing the RPG features is that the system must be modular enough that it should be fully possible to choose away any and all parts of the RPG system at the design level and without changing the source code. In other words: Using the same game engine, it must be possible to make a vanilla gamebook with no features beyond one paragraph leading to the next, as well as gamebooks with full RPG game mechanics.

SKALD has rogue-like features

In addition to the rpg game-mechanics it’s also important for me that the SKALD game engine allows for certain rogue-like features. In particular I want to have the ability to add procedurally generated content to the game. This, combined with an underlying RPG system will make it possible to essentially write a text based rogue-like. At the same time, just as with the RPG system, it must be fully possible to write a gamebook without using a single rogue-like feature.

Why add rogue-like features? Well, my reasoning thus far is that this will make it a lot easier to add to the runtime of the game. I have a vision of the narrative of the gamebooks being superimposed on relatively open worlds and thus adding a lot more replay value to the game beyond just going through the main plot over and over.

Again: This is an optional feature I’ll add to give myself the full breadth tools to tell the kinds of stories I would like.

Engineered with Sustainability in Mind

Not only is SKALD a labor of love, but it will also serve as a tool I could potentially be relying on for years to come. In the long term, it might even be adopted by other game developers interested in working within the gamebook genre.

As with all code that sees long-term use, there’s a lot to be gained by “measuring twice and cutting once“. In other words I intend to take my time and keep the source code architecturally sound, compact and well documented.

The notion of strictly separating the logic from the data and making all aspects of the game itself modifiable via the data is the prime tenet I’m currently building the engine around.

SKALD as a Brand

Lastly, it would be pretty neat if the SKALD system and the game engine in particular becomes a recognizable brand. It would be awesome to see “Powered by the SKALD engine” and “SKALD Roleplaying System presents:”  on games!

 

That’s it so far. Thanks for reading and be sure to keep following the SKALD devlog! In the meantime, follow Scape-IT and SKALD on Twitter for all things RPG!