The production of simulation games requires the development or acquisition of a suitable game engine that is configurable to produce specific simulations. Presented below are the key features and functions that need to be developed in the game engine.
- Events
Events in the context of the game engine should not be confused with events as we generally mean the term in everyday life. In the context of the game engine, events are quantitative data that are either decided upon by the game designers, system administrators, game players or automatically generated by the system either as a random occurrence or in response to other events. A close examination of the previous statement means that events are either input values (eg player decisions) or output values (eg resultant market share) or more typically both (eg player decides on increasing the sales budget which is an input by the player and an output to market share – this is just a simplistic example to illustrate the point).
The following scenario illustrates the mechanism of events. The actions decribed in the diagram occurs during one move or turn during the game. The player decides to increase his advertising budget, which is an event that carries a numerical value. The game has been configured by the games designer to carry the following dependencies – increasing the advertising budget immediately increases costs and reduces profitability in the short run, however it also increases market share which may increase sales and profitability in the long run. Thus the events described by diagram includes the following although various impacts are felt by each of the events during different time frames:
- Advertising Budget
- Costs
- Profits
- Market Share
- Sales
In a realistic game engine with contextual insight, tens of thousands of events may be designed in the engine. Some events may be qualitative measures that are implemented as quantitative values, for example the propensity to spend on luxury goods in a particular income bracket may be defined as a value from 0 (unlikely) to 10 (certainty) with varying levels of propensity in between.
- Dependencies
Dependencies refer to the cause and effect linkages between events. In the above example, there is are dependencies between
- Advertising Budget (cause) to Costs (effect)
- Costs (cause) to Profits (effect)
- Advertising Budget to Market
- Market Share to Sales
- Sales to profits
Dependencies are configured by the game designers during the game design process. In a realistic simulation game, events will have a huge number of dependencies (causes and effects) with one another.
- Clock
The clock is one of the internal mechanism used by game engine to determine the impacts realized through the dependency mechanism. Some impacts are realized immediately, while others may take time for realization and gradual realization over a curve characteristic over a period of time may also be built into the design.
- Dimensions
While events are quantitative values, dimensions are typically non-quantitative values used to further define the particular instance of an event. For example, dimensions applied to the Advertising Budget may be
- Time (when is the increase in Advertising Budget incurred)
- Product (to which product is the Advertising Budget applied)
- Location (in which geographical region is the Advertising Budget applied)
- Media (how is the Advertising Budget used, through which medium is the advertising campaign carried out)
Some dimensions may be quantitative values implemented as quantitative values, for example age of customers may be grouped into bands say pre-schoolers, schoolers, teenagers, young adults, married young adults, etc.
- Fuzzy Logic
Fuzzy logic is best described as what is not fuzzy logic. In a non-fuzzy logic situation a certain cause will definitely produce a certain effect, for example flicking on a light switch will certainly turn on the light. In a fuzzy logic situation, the impact of a certain cause to an effect is determined probabilistically, for example the impact of increasing an advertising budget may increase sales probabilistically. The probability of a cause causing and effect is determined by a probabilistic curve, and the characteristic of the curve is determined by the game designers during the game design phase. The choice of a curve by the game designers may involve a careful analysis of past market data, in which the game may exhibit a better reflect of the real market.
These probabilities are then applied to the game as weightages in the dependencies between events. Thus an event that has 100% weightage has an absolutely certain impact (eg increasing the Advertising Budget has 100% certainty of increasing costs) but dependencies with less than 100% weightage has varying impacts (eg increasing the Advertising Budget has a fuzzy impact on the Market Share.
- Predictive Analytics
Predictive Analytics refer to a special application of the above-mentioned game engine in which the game engine is used to determine the strength of dependencies between defined events. In most cases, all events in the game engine used for predictive analytics are defined as having dependencies with all other events. The game then is started using data that reflects the actual state of players at a real world point of time. Then, various strengths of weightages are applied to the dependencies by almost by trial and error until the model simulates actual real world events that had occurred. The exercise then concludes with dependency weightages that reflect historically how the real world behaves.
The main argument against predictive analytics is that it only predicts market behaviors as long as the market behaves consistently over time. That is to say, since the exercise uses historical data to predict future behaviors, the method only works if the future market behaves as it did in the past. In reality, a predictive analytics is defined using much input from the experience of the analytics designers. For example, the designers may choose historical trends that reflect the dominance of various age groups in the target market, knowing that the age factor is an important determinant in the particular market. This may be applied, for example, in simulating sales of fashion apparels. On the other hand, the analytics designers may place emphasis on other historical market data in say, analysing the engine oil market.
- Constraints
Dependencies are not the only factors that determine the impact of cause events on effect events. Another set of factors that are also taken into account are constraints. For example, increasing the Advertising Budget may increase propensity to be sell, however the Market Size may be a constraint that limits growth potentials.
- Rules
Rules are determined for games to set the dos and don’ts of the game. For example, there may be prohibitions for sellers to sell shares in a bank beyond a certain percentage of shareholding to foreign entities. Rules are introduced into gaming engines to reflect real world dos and don’t’s in the marketplace.
The above required features and functions operate within another set of needs. These include infrastructural requirements to allow games to backed up and restored in case of disasters, responsiveness requirements to allow players to enjoy smooth playing experiences, user-friendly interfaces for game designers, administrators, instructors and players, the ability to roll-back and create what-if scenarios in midst games (a feature not provided by most existing gaming providers), the ability to analyse games and produce reports conducive for post-mortem analysis (a feature not provided by most existing game providers), etc.
Although the additional requirements, are not elaborated in detail, they are nevertheless vital requirements in the gaming eco-system.
Business Thoughts 
Leave a comment