Strategy Game Design Foundations I

If you have never heard of Col. John Boyd, USAF then I would recommend that you follow the link and absorb the info. There is also a great book about Boyd that, despite an ungainly long title, I would highly recommend “Boyd: The Fighter Pilot Who Changed The Art of War.” I don’t know if the accounts of Boyd standing in the doorway of his Pentagon shop yelling “Out There – Business as Usual, In Here- Thunder and Lightning” are apocryphal but it makes for a great story….and fits with the larger than life personality that Boyd must have been.

I bring Boyd up because of one of his seminal contributions to Operational Research/Systems Analysis: The OODA decision cycle.

Basic OODA Conceptualization

This is a basic representation of the OODA loop. I used the concept heavily when I was designing Armageddon Empires and I think that it’s something any strategy game designer would do well to internalize. “Internalize” is a fancy word beloved by military apparatchiks, analysts and academics for putting something in your toolbox that you don’t even think about grabbing…it just occurs naturally that you should use it. Here are some ways strategy game designers can use OODA as a game design tool:

1. Architecting the AI: This is the most obvious one. The concept of a decision cycle easily lends itself to what an AI opponent has to accomplish in a strategy game. For AE I built an OODA cycle that focused on creating and processing goal objects. The Observe phase was dominated by information gathering. Data structures that reflected the AI’s knowledge of the game state provide the raw data. In the Orient phase the raw data is processed and in the Decide phase the goals are generated. Finally a ranked list of goals is processed in the Act phase.

2. Giving the AI personalities: The Orient phase is ideal for this. Boyd had a list of inputs that influence the processing of the raw system observation data. Things like genetic heritage, cultural traditions, and previous experience all figured in the decision process. AI’s can be given the equivalent properties by crafting these type of inputs when it goes about ranking things. I have found that Ranking a list of items is one of the fundamental things that an AI has to do. How the AI ranks a list can yield nice ranges of what is eventually projected onto the game space as “behaviour.”

3. Structuring Game Play Mechanics: This is huge area for application of OODA. When designing the rules set for a strategy game the OODA can help you ask the right questions. What data will players have? What information is gathered from the game board? How can player intentions be manifested into the game state? Should a player have defined pre-existing attributes, skills and perks that influence the Orient phase automatically? Should the Decide phase be constrained or totally free? How is feedback from the Act phase conveyed to the player?

The scope of use for OODA in strategy game design is almost unlimited. Like the analytical framework for security policy that I discussed before it seems so simplistic that it shouldn’t be very important. This couldn’t be farther from the truth. Like the basic concept behind Einstein’s Special Theory of Relativity its simplicity trumpets its beauty and usefulness.

Comments are closed.