Business Simulation Interactions

Here I explore the meta-compositional design of decisions and results in terms of ambiguity and granularity.






Site Map










Home > Meta-Composition Overview > Business Simulation Interactions

This page explores the meta-composition aspects of decisions and results with a separate page exploring the software architecture aspects.

Decisions and results are positioned on the interaction domain with one axis defining their ambiguity and the other defining granularity (detail).

The decision and result interaction domain showing the ambiguity and granularity axes.

Interaction Domain

Decisions and Results

When I keynoted the 2008 ISAGA (International Society for Simulation and Gaming) Conference I suggested that the software had two parts (domains) - the model domain (that consisted of the simulation model and it's associated parameters) and the interaction domain (that consisted of decisions and results.). The interaction domain has two dimensions ambiguity and granularity where ambiguity impacts cognitive processing (learning) and both impact cognitive load (duration).


Meta-compositionally, decisions are designed to elicit thought, discussion and argument and, eventually are entered into the simulation model and assessed. As discussed in the section on task and issue progressions, decisions are often introduced as the simulation progresses.

There are two reasons to incorporate a decision - because it:

  1. elicits the thought and discussion necessary to meet leaning purpose
  2. is necessary for the operation of the simulated business

Eliciting thought and discussion
These decisions are directly associated with learning. When deciding decisions it is important to link each to learning purpose and business success and the results that will help participants unravel the impact of the decision.

Necessary decisions that do not add to learning
These decisions are those that are needed to allow the simulated business to operate in addition to the decisions required for learning. As they add to duration without adding to learning ideally they should be eliminated and this can be done by setting up the simulation parameters so that the decision plays no part, making the decision automatically (using "intelligent" logic). In my Global Operations simulation (a simulation to learn about strategy) as the business grew the company would offer a range of products and these needed to be produced. One approach would be for participants to schedule production (a tactical rather than strategic decisions) or for the simulation to automatically schedule production based on demand, logistics and required inventories). I chose to do the latter and this (together with other actions) meant that Global Operations' duration was only a day rather than the two or more day's duration that is normal for a strategy simulation.

General Point
The disciplined designer will not incorporate a decision for any other reason as this adds to cognitive clutter, increases duration and hence reduces simulation efficiency. Further introducing superfluous decisions can be confusing and hence reduce simulation effectiveness.

When incorporating a decision it is necessary to consider how it impacts cognitive load (duration) and its importance to learning purpose as these help define the need for ambiguity and granularity.


Meta-compositionally, results require analysis to determine how decisions impacted them and identify strengths, weaknesses, opportunities and threats and provide a lead into to planning the next simulated period. As discussed in the section on task, issue and viewpoint progressions, results can be introduced as the simulation progresses.

There are several reasons to incorporate a result - because:the result is necessary to:

  1. help identify the impact of decisions
  2. allow participants to assess their business success
  3. to provide information to support decision-making
  4. to help the management of learning and the tutor
  5. help with design and assure quality.

Identifying decision impact
For learning to occur, participants must identify the impact of their decisions and for this to occur the simulation must incorporate results that provide clues.

Assessing success
Business learning is concerned with improving business success and participants will see their purpose as to make their simulated business a success. These results evaluate business performance and assess success and
besides an objective assessment of success, one should consider the motivational, engagement impact of impact of this assessment.

Providing information
This includes information about competitive actions, customer responses, unexpected (ad hoc) events, cost changes, etc.

Managing learning and supporting the tutor
Here results are provided to help a tutor manage learning - identify learning problems and opportunities and provide the information to help challenge and coach. This is central to a Tutor Support System.

Design Help and Quality Assurance
As discussed on the
Design Verification page I discuss the problems and solutions associated with assuring quality. Key to this is having reports to support the design process - to verify and calibrate models and for a complex simulation more than half the reports are used to support design and ensure quality - reports that are unambiguous and probably granular

General Point
The disciplined designer will not incorporate a result for any other reason. When incorporating a result it is necessary to consider how it impacts cognitive load (duration) and its importance to learning purpose as these help define the need for ambiguity and granularity.


Ambiguity defines the amount of cognitive processing needed to make the right decision and the extent to which the result is relevant to the participants or help the tutor. Decisions and results in the real world are ambiguous. In a business simulation the appropriate level of ambiguity is crucial to getting participants to think through decision-making and result analysis and is crucial to learning. Decisions or results must not be so ambiguous so as to make impossible to estimate the probable impact of a decision or make it impossible to understand the implication of a result. Equally, the impact of decisions and results must not be obvious (unambiguous) or results are so obvious that they do not engender thought and discussion.

Decision Ambiguity

The most ambiguous decisions are those where it is most difficult to understand the links between the decision and outcomes (results) - examples of these (below) are price and promotion decisions where it is difficult to see their impact on sales demand. The least ambiguous are those where the outcome is obvious - often these are decisions about production levels (below) or capacity changes. (Note: For an operation's management simulation production may be ambiguous as participants have to manage inventory shortages, staff efficiency, quality issues etc.)

Ambiguity and granularity example

Often decisions can be divided into two groups - those that deliver the desired learning and those that are necessary to allow the simulation to work. This is illustrated above. The price and promotion decisions are central to learning about the product life-cycle. Whereas, production provides a way of checking that participants are forecasting the impact of their decisions and necessary to explore cash flow.

Result Ambiguity

Result ambiguity measures the extent to which impact of decisions on business success is obvious. Thus there are two categories of result ambiguity - analysis ambiguity and success ambiguity.

Analysis Ambiguity
Analysis ambiguity is the extent to which the impact of the decision is obvious. In the results below penetration is highly ambiguous as in Period 1 is 5% good or bad. The 100% market share shows unambiguously that there is no competition. Result ambiguity often change as the simulation progresses. For example, is the inventory (units) is reasonably ambiguous but if instead of 12029 it fell to zero this would show unambiguously that production was too low and sales were being lost.

Example od result ambiguity

Success Ambiguity
Success ambiguity is the extent to which it is obvious that the simulated business is successful and depends on the complexity the success metric.

Above there a just two, separate success measures - cumulative profit and cash. Cumulative profit shows how successful the company is at generating profit and cash shows bankruptcy risk. In contrast, a Total Enterprise Simulation selling multiple products or into multiple markets with Profit Centre reports, Profit and Loss (Income Statement) and a Balance Sheet will have multiple success measures (growth, market share, profit to sales, ROI, Capital Gearing (Leverage) etc.

Multifaceted success metrics ensure a rich discussion about business purpose and, often, lead to different goals for participating teams but this adds to duration. In contrast, simple, unambiguous, metrics (as above) do not need discussion and hence are suitable for short duration simulations.

Ambiguity Drivers

Ambiguity is impacted thus:

Natural Ambiguity
As illustrated above depending on the simulation and the real-world situation both decisions and results have a natural ambiguity.

Range and Importance
In the decision ambiguity above sales demand is impacted by both price and promotion and this increases the ambiguity of sales demand as participants must unravel the extent to which each contribute to demand. My PriceWize simulation had a single prime performance measure - profit contribution and this meant that result ambiguity was low. In contrast my Challenge Series of Business Acumen simulations have multiple prime performance measures (operating profit, earnings, cash flow, ROI, market share and liquidity) and this meant that assessing business performance was highly ambiguous.

Temporal-Topical Progressions
The way the simulation progresses over time impacts ambiguity. In particular, economic and business progressions tend to add to ambiguity and task, issue and viewpoint progressions can be used to reduce ambiguity. As the market is penetrated (business progression) in my Product Launch simulation price sensitivities change and this keeps price ambiguity high. In contrast sales variation, demand growth and seasonality (economic progression) means that for my Operations simulation the production decision remains ambiguous. Reports can be introduced (viewpoint progression) to help explain the impact of decisions and hence reduce ambiguity as participants must separate out the impacts of demand growth, seasonality and random fluctuation. Additional performance measures can be progressively introduced to explore new issues (issue progression) and this was done in my PriceWize simulation - initially the only concern was using price to maximise sector profit contribution but then idle staff costs were introduced meaning that price not only had to maximise sector profit but also using it to reduce idle staff costs and through this maximise overall profits. Also, my PriceWize simulation introduced value change decisions (task progression) affecting price ambiguity.

Decision-Model-Result Linkages
The complexity of the linkages between decisions-the-model and results impacts both decision and result ambiguity. My PriceWize simulation has uses price to assess demand and through this sales and profits. In contrast, my Service Challenge simulation takes price, promotion and perceived service quality to determine demand from existing customers and new customers. Demand is then matched to capacity across three market sectors and two resources to determine actual sales,. Actual sales is then used to calculate revenues, costs, profits and cash use. - a reasonably high ambiguity.

Participants' Manual, Cognitive Prompts and Reflection Triggers
If appropriate, textual information is used to reduce ambiguity and there are three ways to do this. Commonly, I provide business history or market research information in the participants' manual. My Service Challenge simulation provides three year's trading history with decisions carefully chosen to provide "clues"
about how decisions impact results. My Product Launch simulation involves the launch of a unique product and to help reduce ambiguity a test market provides information about reasonable prices, promotion and production.

Simulation Type
Whether the simulation is interactive between teams or whether teams operate independently impact's ambiguity. Teams interact besides the ambiguity built into the simulation there is also the ambiguity of the impact of the other teams' decisions.

Ambiguity Design Issues

As illustrated below ambiguity must be decided purposefully based on learning needs and balancing the cognition needed against the cognitive load rather than based on real-world ambiguity. In the context of learning efficiency and effectiveness one should link the ambiguity level to learning importance. As shown below it is likely to be necessary to reduce ambiguity in the simulated world from that of the real world. Real world ambiguity is illustrated by this quote attributed to John Wanamaker "Half the money I spend on advertising is wasted; the trouble is I don't know which half" but for learning to occur, eventually, learners must know which half.

Ambiguity Spectrum


Granularity defines the level detail for decisions and results and impacts the amount of cognitive processing required to settle on the decision and analyse the result and needs to be determined to balance this with necessary cognition.

Decision Granularity

Decision Granularity is high when the decision is a number (top left) as the range of options is high. Decision Granularity is reduced as the number of decision options reduce through multiple choice decisions to binary (yes/no) decisions.

To the top left are numeric decisions - the most granular of these is Inventory Purchases where a four figure decision can be entered and the least granular Percent Markup where the decision is to the nearest percent. Below this some multiple choice decisions are shown - illustrating low granularity.

Decision Granularity Spectrum

Result Granularity

Result Granularity defines the amount of processing done by the participants compared with the amount of processing done by the software. So, basic accounting results are highly granular and require significant work done by the participants. Results become less granular as they are used to provide financial measures (such as ROI) and less granular still as they are presented as trends or graphs.

The Profit Center report (left) is granular but even here granularity was restricted by the size of the numbers. Even so, to decide whether the Profit Centre is adequately profitable it is necessary to process the results further (by dividing Contribution by Net Assets to determining ROI). Even when this is done one must decide whether the ROI is good, bad or OK. Also, the Profit Centre Analysis report is for a single period and so does not show trends in revenue, contribution, asset needs and through this cash flow and growth.

The Dash Board has low granularity as it shows results and indicates whether these are good, OK or bad eliminating the need for participants to decide what are the right levels of Productivity and Capacity Use.

Result Granularity Spectrum

A particular issue with result granularity is the number of reports produced as this can lead to cognitive overload and analysis paralysis. However, to an extent this can be overcome with appropriate task, issue and viewpoint progressions.

Granularity Design Issues

Deciding granularity involves a purposeful analysis of the importance of the decision and result and the amount of cognitive processing required (learning effectiveness). The amount of time that can be budgeted to the making the decision or analysing the result) is key to meeting duration constraints (learning efficiency) and this impacts deciding granularity.

2015 Jeremy J. S. B. Hall

Most recent update: 29/07/15
Hall Marketing, Studio 11, Colman's Wharf, 45 Morris Road, London E14 6PA, ENGLAND
Phone +44 (0)20 7537 2982 E-mail