An Archeology of Strategy Games

Don’t mind the pretentious sounding title. This is going to be a nuts and bolts entry that attempts to lay down some of the foundations for thinking about strategy game design….at least how I think about strategy game design. Being an amateur systems engineer and a formally trained national security policy analyst I like to create diagrams and “analytical frameworks.” Here is some of my best work:

Strategy Game System Chart

This homespun illustration points out the three major components of the strategy game system. The players (human and AI) are depicted outside the system box. They function as the inputs to the system. They formulate a strategy and then press keys and click the mouse to input their commands. Their commands are interpreted by a rule set. The rule set is far more than a command interpreter however. It also specifies how data within the game state space is processed. The data is represented in the diagram by the box with the Map/Agents label.

Armageddon Empires is built on this model. There are basically two types of objects in the game. Game data objects that have some specific set of attributes associated with them and must be able to be saved. The structure is hierarchical as well. A Player Controller Object contains Player Objects. Player Objects contain an Army Controller Object which in turn contains Army Objects. An object for you non-programmers is just a collection of data elements and operations (functions) that all work together. For example a Car object implemented in software would have data for its number of passengers, max speed etc. and the operations would be start, accelerate, stop. The diagram above doesn’t always translate perfectly however into computer code. I’m not a great programmer so a lot of the data objects have rules built into them. The other type of object I use is an interface object. These objects display menus, create the map, and retrieve and store the data of the Game Objects. The rules of the game emerge from their implementation. The battle module in Armageddon Empires is a perfect example. The way the cards are presented, the structure of taking turns attacking, and the presentation and processing of the die rolls are all accomplished by interface objects.

Notice the little Victory Conditions gauge coming out of the top of the game system box. The idea is that the state of the data is measured and some arrangements are better than others for the players. You win a game of Armageddon Empires by creating a game state where your opponents no longer have strongholds on the map. It’s pretty funny to think about what you are really doing when you play a game of Civilization for example (or any game for that matter). You spend hours staring at a monitor pressing buttons and clicking a mouse to arrange 1′s and 0′s into a favorable pattern that is massaged in a microprocessor and stored long term on a hard disk. You do all this to achieve a victory condition that can be as mundane as a screen that tells you something like “You have crushed your enemies and now rule the wasteland.” You don’t even get a food pellet!

The fun (or burden) of being a strategy game designer is that you are responsible for creating a fully functional data space and rule set. In other words you have to build a universe. How well the rules perform on the data set and how “fun” it is to create a winning pattern of 1′s and 0′s is the standard by which your creation will be judged. I’ll leave an exegesis on “fun” to the pit fighters at QT3 but I know it when I see it.

Contrast the game designer’s role with that of the national security policy analyst’s role. The analyst is presented with the rule set and game state structure as a given. The world is as it is…anarchic, organized around nation states and unsettled by the recent emergence of non-state actors. This next diagram explores some more of the structure that resides in the inputs. It’s the domain of the general, admiral and national security policy maker.

Behold the Agents and Actors

Those squiggly things are brains. They make plans based on what they know about the rules and data and what they think they know about other brains. The text next to the bottom brain is what is known as an analytical framework. You are looking at an abridged version of the best thing that I learned at Harvard’s JFK School of Government. It looks rather simple but it is in reality a very effective tool. The basic International Security Policy class (ISP 201) was focused exclusively around it…an entire semester. It was taught to me by Ambassador Robert Blackwill who was recently ambassador to India and also played a key role in the reunification of Germany. I’ll examine it in more detail in a future entry. The reason I want to take a look at it is because I have found it very helpful in using it as a tool to not only understand current events but also to create rulesets and data spaces. Analyzing in a formal way the manner in which players will formulate strategy to interact with your design is extremely beneficial. It’s also very useful when initially conceptualizing the AI before you have done any observation of how humans will actually play the game. More later…

Comments are closed.