Chapter 3

Problem-Solving Strategies
For Managing Uncertainty

It is not possible to solve a problem using the same thinking that created it.
– Albert Einstein, theoretical physicist.

Confrontation Modes

Previous chapters have identified a number of factors which influence a project’s

approach to uncertainty:

• novelty – projects which venture into new territory face greater unknown
challenges.
• complexity – large numbers of co-dependent tasks or events introduce
greater potential for uncertainty.
• affordability – the cost of managing uncertainty has to be balanced against
the threat. It may be uneconomic to tackle some areas of uncertainty.
• structure – the organization of the project into tasks and the choice of
delivery approach determine how uncertainty develops through different
project stages.
• capacity – projects have a higher capacity for uncertainty if the consequences
of failure are judged to be acceptable. For this to be justified, the rewards of
success must also be high. High capacity projects gamble that the benefits
Copyright © 2016. Taylor & Francis Group. All rights reserved.

of success are worth the considerable uncertainties that must be overcome.

Taking these factors into account, a project manager must then decide how best
to confront uncertainty. For example, a risk averse project with a low capacity for
uncertainty will focus on suppressing as many sources of uncertainty as possible.
But it will also need to be responsive to new threats and find ways to avoid or
detour around these unknowns.

Figure 3.1 shows four possible modes for confronting uncertainty.

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 38 Managing Project Uncertainty

1. Suppress – take pre-emptive steps to reduce overall levels of uncertainty

before problems occur.
2. Adapt – try to contain uncertainty to acceptable levels, deal with unexpected
outcomes as they arise, and remain focused on the objectives.
3. Detour – find an alternative way to reach the objectives which avoids the
uncertainty. By following a different path, any unexpected outcomes that
do arise are either irrelevant or have less impact on the project.
4. Reorient – if unavoidable uncertainty means the project goals are unlikely
to be achieved, find acceptable alternatives (or compromise).

pre-emptive steps to
reduce uncertainty

start planned path

SUPPRESS point objectives

adaptive planning

start planned path

ADAPT point objectives

lan
rep
start
DETOUR point uncertainty objectives

recognize
uncertainties
Copyright © 2016. Taylor & Francis Group. All rights reserved.

insurmountable
uncertainty

start
REORIENT point objectives

recover and
replan varian
t plan new
objectives

Figure 3.1 Four possible modes for confronting uncertainty

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 problem-solving strategies for managing uncertainty 39

Suppression

A strategy of suppressing uncertainty is the most commonly adopted of the four

modes. Pre-emptive tactics reduce the uncertainty at source, leaving the project to
proceed smoothly against the optimal plan. Avoiding uncertainty by suppressing
it sounds like a safe bet – and it is, providing it can be done cost-effectively.
However, for novel or highly complex projects, particularly those with many co-
dependencies, it may be too difficult or costly to suppress all possible areas of
uncertainty.

Adaptation

By adapting, a project manager tolerates a working level of uncertainty but is

prepared to act swiftly to limit the most damaging aspects of any unexpected events.
This is a highly pragmatic approach. It requires agile and flexible management
processes which can firstly detect emerging issues in their infancy and secondly,
deal with them swiftly and decisively. For example, imagine a yacht sailing in
strong winds. The helmsman cannot predict the strength of sudden gusts or the
direction in which the boat will be deflected, but by making frequent and rapid
tiller adjustments, the boat continues to travel in an approximately straight line
towards its destination.

Detouring

Given the choice, we should like to detour around all areas of uncertainty. Avoiding
the source of uncertainty means that the consequences (that is, the unexpected
outcomes) are no longer relevant to the project. Thus there is no need to take costly
precautions to resolve unknowns or deal with their repercussions.

Unfortunately, detouring around uncertainty is hard to achieve, for two reasons.

Firstly, many sources of uncertainty are simply unavoidable, or the avoidance
measures are too costly. Consider the earlier example of a subcontractor who may
Copyright © 2016. Taylor & Francis Group. All rights reserved.

be incapable of delivering a critical input on time. We could detour around this

uncertainty by dismissing the subcontractor in favour of a competitor who can
provide a better service. This will mean cancelling existing contracts, researching
the marketplace and renegotiating commercial terms with an alternative supplier
– all time-consuming and potentially costly activities – and with the risk of being
no better off with the alternative supplier.

Secondly, detouring only works for quantifiable uncertainty (the ‘known

unknowns’). Unfathomable issues often loom up out of the mist of uncertainty too
fast to permit a detour.

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 40 Managing Project Uncertainty

Reorientation .

Our final option is reorientation. This is a more dramatic form of detour where
we aim for a modified set of objectives in the face of insurmountable uncertainty.
Highly novel projects sometimes require this. To plough on in the face of extreme
uncertainty risks total failure. The only alternative is to redefine the goals, that is,
reorient the project in a way that negates the worst of the uncertainty.

This is not a tactic for the faint-hearted. Convincing the client that a project cannot
be delivered as originally conceived is no easy task. But it is worth asking the
question, ‘Is it better to deliver something different (but equivalent), than nothing
at all?’

Table 3.1 More than one way to deal with uncertainty

Confrontation mode Implications for the project

Suppression Uncertainty is removed or reduced by tackling its root
causes. By converting vague assumptions and poorly
understood issues into tangible, measurable problems,
management processes can take these factors into
account. This means:
• identifying the sources of uncertainty;
• accurately forecast future scenarios;
• analysing potential threats;
• converting any area of uncertainty into one or more
specific problems;
• developing a tactical plan to resolve the problems.
Adaptation A certain level of uncertainty is tolerated, typically
because there are insufficient resources to suppress
it, or such actions would prove too costly. The threat is
contained by closely monitoring key uncertainties and
Copyright © 2016. Taylor & Francis Group. All rights reserved.

reacting to early warning signs. This means:

• having good information-gathering and early warning
systems in place;
• understanding (and prioritizing) the key objectives in
order to stay focused on the right things if the plan
has to change;
• acting rapidly and decisively;
• continually steering the project towards its goals.

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 problem-solving strategies for managing uncertainty 41

Table 3.1 Concluded

Detour Is there an alternative path to the same objectives? This

may not be the optimal route planned at the start of the
project, but becomes viable once major uncertainties
are identified. Be careful about avoiding one set of
uncertainties only to replace them with different ones.
This means:
• reaching a clear understanding of the project
objectives which is shared by all stakeholders;
• being creative in identifying planning options;
• evaluating and comparing the risks and benefits of
alternative approaches;
• seizing the initiative when better opportunities present
themselves.
Reorient Does the level of uncertainty represent too great a risk?
Redefining the objectives may appear to be the case of
thinking the unthinkable, particularly if they are written
into commercial agreements. However, faced with the
choice between project failure or successfully delivering
against alternative goals is a viable option. This means:
• understanding the nature of the threat (for example,
being realistic if confronting insurmountable
uncertainty);
• being honest about chances of success and failure;
• keeping an open mind on redefining objectives;
• being persuasive in seeking the agreement of
stakeholders to reorient the project.

Exposing the Underlying Problems Caused by

Uncertainty
Copyright © 2016. Taylor & Francis Group. All rights reserved.

Problem-solving lies at the heart of an effective management strategy, no matter

which confrontation mode is chosen. If we opt to suppress or detour around
uncertainty, the problems are centred around forecasting: it is necessary to
understand which areas of the project are most vulnerable in order to focus on
suppressing or avoiding the appropriate areas of uncertainty. An adaption strategy
means using problem-solving skills to deal with unexpected events as they occur,
and reorientation is a special case where the problems are severe enough to consider
redefining the goals. In all of these situations, problem-solving is a vital tool. The
rest of this chapter will examine effective methods of problem-solving and look at
the particular difficulties arising when many of the key variables are uncertain.

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 42 Managing Project Uncertainty

Problem-framing and problem-solving

Problem-solving breaks down into two parts. Firstly, problem-framing transforms

a particular area of uncertainty into a set of specific problems. By so doing, we can
better understand the nature and magnitude of the threat posed by the unknown. It
lets us answer important questions: What are the implications of this uncertainty
(that is, the range of possible outcomes)? What effect might they have on the
project? Is this an acceptable risk or a show-stopper?

Common experience tells us that we can’t find answers until we properly understand
the question. Problem-framing is the process for doing this. We take a particular
area of uncertainty (say, doubts over how a product will integrate into a client’s
infrastructure) and tease out the set of problems arising from this uncertainty. This
means asking sensible questions and challenging any unsubstantiated assumptions:
How do we find out if there are fundamental incompatibilities? What interface
design methods should be used? How should we test the integration? By defining
the questions that need addressing, we have taken the first step towards managing
the uncertainties.

Secondly, solution-finding techniques provide a way to efficiently resolve the

uncertainty. This means choosing from many possible solutions the one which
delivers the greatest value in terms of meeting the project objectives. The solution
may be to act on the source of the uncertainty (that is, find ways to effectively
suppress or detour around it) or, in response to unexpected events, find an outcome
which lets the project adapt to the uncertainty.

Problems and uncertainty are just two sides of the same coin: a problem is a
restatement of a particular uncertainty. The point is, by taking an area of uncertainty
and restating it as a problem (or set of problems) – the process of problem-framing
– we can then apply classical problem-solving techniques to reduce the level of
uncertainty.
Copyright © 2016. Taylor & Francis Group. All rights reserved.

Getting to the Root of Uncertainty

Before we can look in detail at problem-solving techniques, there are three kinds
of traps to be wary of:

1. faulty pattern recognition;

2. confusing symptomatic problems with root problems;
3. vulnerability from independent uncertainties.

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 problem-solving strategies for managing uncertainty 43

Faulty pattern recognition

The human brain, although superb at pattern recognition, can be easily fooled.
Because of superficial similarities, it is not uncommon to mistake a novel situation
for a problem we have previously encountered. It is an instinctive reaction to
ask, ‘Where have I seen this before? What does this remind me of?’ Drawing on
experience and learning from previous successes and failures is undoubtedly a
good thing, but misinterpreting a fundamentally different situation has dangerous
consequences. We can end up trying to solve a different problem to the one we are
actually confronted with.

Objectivity is therefore important. Ask yourself:

• What are the key differences between this situation and similar problems I
have encountered in the past?
• Are these significant differences, for example, ones that act on the problem
differently or at a different tempo?
• Is the scope of the problem different? Are there a larger number of variables
at work?
• How timely is the previous experience? Circumstances change, methods
evolve, and commonly accepted wisdom is challenged. Has your previous
solution to this kind of problem become outdated?

Symptomatic and root problems

Have you ever solved a problem only to find essentially the same issue cropping up
again a short while later? This typically happens when confronting a symptomatic
problem. Symptomatic problems appear to be discrete and self-contained but
actually conceal a deeper underlying problem. Because there are bigger issues
at work than those under consideration, not surprisingly, any solution to a
symptomatic problem tends to be ineffective or short-lived because it doesn’t deal
with the underlying or root problem.
Copyright © 2016. Taylor & Francis Group. All rights reserved.

A doctor knows that it is more effective to treat the underlying cause of illness
than merely address the symptoms. The underlying illness (or root problem) can
be determined by using the symptoms as clues in the diagnosis. Tackling project
uncertainty requires the same approach: the symptoms of uncertainty hold the
clues. Why is this task running behind schedule? There could be many root causes:
lack of resources, a poorly skilled team, unforeseen dependencies or perhaps poor
planning and estimation.

When we get involved in the details of framing the problems arising from uncertainty,
we sometimes miss the bigger picture and mistake symptomatic problems for the
root problem. By focusing too closely on gathering missing information, so much

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 44 Managing Project Uncertainty

time and energy is spent on converting unknowns to knowns – through research,

knowledge discovery, modelling or trial-and-error – that we lose sight of the real
implications of the uncertainty.

When framing a problem it should be remembered that the nature of the parts are
not always representative of the whole – for example, a collection of springs and
cogwheels doesn’t tell us much about how a clock functions. The root problem is
not always obvious from a cursory examination of the symptomatic problems.

Breaking a problem down into smaller pieces is often seen as a good strategy,
since smaller problems are easier to deal with. However, by only addressing
symptomatic problems, the root problem – which lies at the heart of the uncertainty
– is untouched. As Figure 3.2 shows, not all problems are fundamentally divisible.
Even though each symptomatic problem has been addressed, some indivisible
aspect of the root problem remains. In this case, the sum of the parts is less than
the total.

Divisible problem

�
� � Solving all of the
constituent parts solves
the main problem.

� �

Indivisible problem
�
Copyright © 2016. Taylor & Francis Group. All rights reserved.

� root � Solving all of the

problem constituent parts leaves
� behind an unresolved
root problem.

� �

Figure 3.2 Complex problems are often more than the sum of their parts.
Breaking this kind of problem into smaller units and solving each
one individually will leave behind an unresolved root problem.

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 problem-solving strategies for managing uncertainty 45

Analysing the root problem

Figure 3.3 shows the steps needed to arrive at the root problem. First, collect as
many symptomatic problems as possible. (A major area of uncertainty will likely
give rise to quite a few.) It is important not to be too judgmental about what is
relevant or important; working from preconceived ideas carries the risk of missing
important implications. Avoid the temptation to filter out problems which appear
unrelated; crucial dependencies or relationships can be missed.

Next, comes the step of diagnosis. Are there common themes which point to
an underlying problem? Is there a pattern of association or causal links (that
is, a preordained sequence or behaviour pattern)? From this diagnosis it should
be possible to derive a clearer picture of the root problem at the heart of the
uncertainty.

The final step is an important one: test the root problem to see if it creates the same
set of observed symptoms. If not, it may not be the right root problem and a further
iteration of the process is needed.

1
collect
symptomatic
problems

4 2
validate root
problem diagnose
underlying
Copyright © 2016. Taylor & Francis Group. All rights reserved.

against
observation causes

3
derive root
problems

Figure 3.3 A four-step process for analysing root problems

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 46 Managing Project Uncertainty

Objectivity and insight

It is human nature to look for evidence which supports a preconceived notion and
to explain away items which don’t accord with the model. However, symptomatic
problems which don’t quite fit with the root problem should never be ignored.
Objectivity is vital: a symptom at odds with the model is indicating that there are
still more dimensions to the root problem to be uncovered.

Let’s look at an example: At the start of a project to develop a management

reporting system, the team face a key uncertainty – which database product to
base the bespoke system on? A transactional database product could deliver a
fast, enterprise-wide service to a large number of users but might sacrifice some
reporting flexibility. On the other hand, a business intelligence product would
allow users greater freedom to create ad hoc queries and slice-and-dice the data to
suit their own purposes. However, the client’s requirements are vague on which of
these scenarios will predominate.

This one crucial uncertainty gives rise to all sorts of symptomatic problems. What
trade-off should be made between speed and functionality? What size user base
should be supported? What growth in data volumes should be planned for?

Table 3.2 Guidelines for framing problems arising from uncertainty

Dig deeper Don’t accept evidence of a problem at face value.

Look for the root Is there a more fundamental problem which explains and
problem summarizes a set of symptomatic problems?
Déjà vu Many uncertainties look like something encountered
before. Are these superficial similarities, in which case
previous solutions may not work?
Beware of Breaking a problem down into bite-size chunks may still
simplification leave behind the root problem (see Figure 3.2).
Copyright © 2016. Taylor & Francis Group. All rights reserved.

Objectiveness Leave behind preconceived ideas on the nature of the

problem. Be guided by what the observable evidence
says.
Mutually exclusive, The global consultancy firm McKinsey coined the
collectively acronym MECE (Rasiel 1998) to check that the facts
exhaustive (MECE) describing a problem are mutually exclusive, (that is,
each fact is stated only once, without duplication) and
collectively exhaustive (that is, all the pertinent issues
have been considered with equal thoroughness). This is
vital for successful problem-framing.
Test the problem Does the root problem account for all the observed
statement evidence? (If not, more work is needed.)

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 problem-solving strategies for managing uncertainty 47

It is conceivable that each of these problems can be systematically addressed and

the inherent uncertainty reduced to manageable levels, resulting in a set of design
decisions which represent the best possible compromise. But this misses the root
problem. Why isn’t the client able to provide firm requirements? Is it possible
that the client doesn’t really know what is needed? If true, the solution may be
to implement none of the possible database solutions – what the client needs is
more analysis or a period of prototyping to enable the requirements to be further
developed.

Vulnerability from independent uncertainties

Launching a new drug onto the market typically costs around €750m and takes 12
years of exhaustive clinical trials before its efficacy and safety can be established
beyond reasonable doubt. Collectively, pharmaceutical and biotechnology firms
annually invest in the region of €70b in research and development. Much of the
cost comes from conducting extensive trials which are designed not only to verify
the drug’s effectiveness, but to detect complications if different drugs are taken
in combination, or where side-effects may be caused by certain environmental
factors. Unfortunately, problems can still occur and a surprising number of drugs
are withdrawn on safety grounds, despite receiving full approval for clinical use.

This is a classic example of vulnerability arising from independent uncertainties.

What happens is that unconnected events combine in unforeseen ways to expose
a latent flaw or weakness in the system. Individually, each area of uncertainty is
small and inconsequential; it is only when circumstances cause these elements to
interact in an unlikely way, that a major problem arises.

James Reason, Professor Emeritus (and formerly Professor of Psychology) at

the University of Manchester, has studied this phenomenon extensively (Reason
2000). His ‘Swiss cheese’ model shows how this kind of uncertainty can lead
to a major failure. Any complex undertaking such as a large-scale project or the
introduction of a new drug has built-in barriers to guard against certain types
Copyright © 2016. Taylor & Francis Group. All rights reserved.

of failure. These are the safety procedures, good working practices and proven
management processes developed over the years to avoid preventable problems.
However, these barriers are peppered with tiny holes (like a Swiss cheese) arising
from the little uncertainties that can never be totally eliminated: the risks we
haven’t foreseen, the variables we hadn’t considered, etc. Normally, these don’t
matter too much. These uncertainties fall within acceptable tolerances, and if one
defensive barrier is breached, there are other lines of defence.

But suppose the holes in each successive barrier just happen to line up? The chance
alignment of a whole series of minor uncertainties suddenly combines to cause a
major (and potentially disastrous) outcome.

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 48 Managing Project Uncertainty

The difficulty is, there is no underlying root problem to be discovered. Our only
chance of detecting such a vulnerability is to work back from a disaster scenario
and identify the circumstances (no matter how unlikely) which could give rise to
it. By understanding the causal factors, we can identify which areas of uncertainty
require attention.

This is not unlike brainstorming around how to achieve the project’s objectives,
except the aim is to identify (and then avoid) negative project outcomes. It requires
‘backward thinking’ – starting with an end result (in this case, a highly undesirable
one) and thinking backwards to identify the conditions which must exist to create
it. We will explore this further in the anticipation strategies of Chapter 5.

Framing the Underlying Problem

The main benefit of framing a problem is that it provides a structured way of

gathering missing information and, equally importantly, of discovering fresh
unknowns. The act of framing a problem drives us to develop a sound understanding
of all sorts of different aspects of the underlying uncertainty: how far it extends,
what implications arise from it, how quickly the situation is likely to change – in
fact, all the characteristics of uncertainty we have already identified. It may require
some kind of representational model to be built so that different aspects of the
problem can be examined in different scenarios.

It enables us to approach uncertainty from a different angle. Instead of beginning

with one or more sources of uncertainty and attempting to suppress or adapt to
them, we work outwards from the centre of the problem. By discovering the
root problem, we determine which key uncertainties must first be tackled. Going
through the steps of problem-framing often reveals uncertainties we didn’t know
existed and is an important way of moving ‘unknown unknowns’ across into the
‘known unknowns’ quadrant.
Copyright © 2016. Taylor & Francis Group. All rights reserved.

Problems often aren’t what they seem at first sight, particularly when information
is incomplete. Their implications may be more far-reaching than we realize or
have an impact that is felt in an unexpected way. We therefore need a formal
method of formulating and evaluating the possible solutions. One such approach
is shown in Figure 3.4.

The essence of framing the problem is to capture as much information as possible

about the following components:

• controllable variables;
• uncontrolled variables;
• constraints acting on the variables;

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 problem-solving strategies for managing uncertainty 49

• relationships between variables;

• evaluation criteria (that is, things which determine what a successful
solution looks like).

Imagine the following problem scenario. Whilst driving along a minor road, a
motorist comes up behind a large tractor chugging slowly along. The driver is
already late for a meeting and anxious not to be delayed further, but unfamiliarity
with the road means the driver can’t be certain if there are safe passing places
ahead, whether there are short-cuts or, indeed, if it is likely the tractor will reach
its destination shortly. Using this simple example, we will examine each of the
problem components as an exercise in framing the problem.

do research to find
missing data

MODEL apply creative

thinking
constraints

uncontrolled evaluate
variables

possible
relationships outcomes
select
of the
constraints model
controllable
variables
validate
Copyright © 2016. Taylor & Francis Group. All rights reserved.

solution
undesirable
evaluation
implications
criteria

intrinsic value extrinsic value

Framing the problem Finding the solution

Figure 3.4 A two-stage approach to problem-solving. Firstly, the nature of the

core uncertainty is established by ‘framing’ the problem. Secondly,
candidate solutions are evaluated and the one which represents the
best value (within the context of the project objectives) is selected.

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 50 Managing Project Uncertainty

Controllable variables

Controllable variables are the things which we, as the principle decision-maker,
can alter or influence. In arriving at a solution, we are effectively deciding which
controllable variables to adjust, and by how much. This set of variables defines
our options. Although other solutions exist which aren’t governed by controllable
variables, these are by definition beyond our influence and therefore cannot form
part of our decision-making process. In other words, we can only make decisions
about things we have some element of control over.

In the example above, both the speed and direction of our vehicle are controllable.
By manipulating these, one possible solution is for the driver to overtake the
tractor. Another solution, arrived at by controlling the same variables in a different
way, is for the driver to wait patiently behind the tractor. The most satisfactory
solution will depend on the relative importance of the objectives – to make the
meeting on time, or to travel safely above all else.

Just because we can identify a controllable variable doesn’t mean it is relevant to

the problem. We must justify why a controllable variable is worthy of consideration,
or the problem will quickly become lost in a swarm of irrelevant variables. There
are many other aspects of the vehicle under the driver’s control (for example, the
amount by which the window is wound down) but none of these have any direct
bearing on the overtaking problem.

Uncontrolled variables

Uncontrolled variables include all the things that we (the decision-maker)

cannot control but which nevertheless affect the possible outcomes. Some may
be controlled by others who are partially within the project manager’s sphere of
influence, (for example, a project stakeholder). Other factors may be inherently
uncontrollable.
Copyright © 2016. Taylor & Francis Group. All rights reserved.

In our example, some of the key uncontrolled variables are:

• The geography of the road. (Are there places wide enough for
overtaking?)
• Forward visibility. (How far can the driver look ahead to see if it is safe to
overtake?)
• The presence of oncoming traffic. (Are there gaps which would allow safe
overtaking?)

Although we can’t directly affect uncontrolled variables, they are still an important
consideration. They define the boundary of what is within our control and what
lies outside.

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 problem-solving strategies for managing uncertainty 51

Constraints acting on the variables

Both types of problem variables, controllable and uncontrolled, have constraints.

There are limits to the speed and manoeuvrability of the car, the width of the
road, the sharpness of bends, the maximum speed of oncoming vehicles and the
driver’s fastest reaction time. Knowing what these constraints are fine-tunes the
representational model of the problem. Within these constraints, the combination
of possible values of the variables creates a set of solutions we can choose from.

Relationships between variables

Complex relationships may exist between the problem variables. For example, the
greater the driver’s speed, the swifter the overtaking manoeuvre but less time is
available to avoid oncoming traffic, so the driver’s reactions need to be quicker,
etc. We also need to understand which variables have the greatest influence over
the outcome. There are three main types of relationship:

1. Strong causal relationship. A strong causal relationship exists where

there is a chain of predictable and repeatable events. (‘Strong’ refers to the
deterministic nature of the chain of events). Strong causal relationships are
comparatively rare outside of natural and scientific phenomena. A beaker of
water placed in a freezer will always turn into a block of ice. Therefore, we
can say there is a strong causal relationship between the initial event (water
placed in a freezer) and the outcome (ice), due to the physical properties
of water. The advantage of causal relationships is that if we understand the
underlying principles of the relationship, we can make accurate predictions
of future outcomes.
2. Weak causal relationship. A weak causal relationship occurs where there
is only a given probability of event A leading to outcome B. For instance,
there is a weak causal link between incubating an egg and a chick being
hatched. B cannot happen unless A has taken place (that is, the egg won’t
hatch unless incubated) but the outcome (hatching) is not certain since
Copyright © 2016. Taylor & Francis Group. All rights reserved.

it is possible that the egg has not been fertilized. The majority of causal
relationships are weak. Our objective is to establish not only the nature of
the relationship (that is, the causal chain of events) but the factors which
determine the probability of each event in the chain occurring.
3. Association. An association exists where we can use one variable to
predict the value of another, but changing the value of one variable doesn’t
necessarily bring about a change in the other. An oft-quoted example is
the association between a person’s height and their weight. We can predict
with reasonable accuracy a person’s weight given their height. However, it
doesn’t follow that reducing a person’s weight will reduce their height, as
might be expected if there was a causal relationship. An association enables
us to predict certain aspects of variables but it won’t necessarily explain

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 52 Managing Project Uncertainty

their values. Mistaking associations for causal relationships is a common

problem, since the differences are often subtle. Firm evidence is needed
before assuming a causal relationship.

Evaluation criteria

The evaluation criteria are a frequently overlooked aspect of problem-solving. How

can we tell if we have arrived at a good solution or if we need to keep looking? It is
impossible to judge unless we have a clear way of evaluating potential solutions.

As always, it is vital to keep the project objectives in mind. These provide a

context for judging the value of a solution. The higher the perceived value, the
more successful the solution.

In the overtaking problem, here are some possible evaluation criteria:

1. How well does it reduce the chance of being delayed?

2. Is it a safe solution?
3. Does it require more data to be gathered (for example, to consult a map for
alternative routes)?
4. Does its success depend on uncontrollable variables?

Depending on the driver’s objectives, different weightings might be assigned to

these criteria. If a major business deal hinges on a timely arrival, criterion 1 is very
important and a solution involving a bold overtaking manoeuvre may score highly.
If the objective is to travel safely above all else, criterion 2 predominates. Being
patient or finding an alternative route may be the optimum solution.

The fourth criterion ensures that the success of the solution doesn’t depend on
external (and unpredictable) factors. For example, the ‘success’ of overtaking on a
blind bend depends on whether or not there is an oncoming vehicle which clearly
isn’t under the driver’s control, and is therefore not an acceptable solution.
Copyright © 2016. Taylor & Francis Group. All rights reserved.

Finding the Solution

Evaluating possible outcomes

The problem-framing process leads to a representational model of the root problem.

This model enables the set of possible outcomes to be explored, each of which has
an inherent value. Outcomes with a high value are desirable; these represent a
good solution to the uncertainty and they accord with the project’s objectives. A
low value outcome is still valid, but isn’t regarded as a good solution.

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 problem-solving strategies for managing uncertainty 53

The value of the outcome can only be judged in relation to the project objectives.
Does this solution move the project closer to its objectives? Is this a cost-effective
way to address the uncertainties around which the problem has been formulated?
A high value outcome will be able to answer yes to both these questions. There are
usually many possible solutions, but only a very small number which deliver high
value outcomes. The art of problem-solving is therefore to arrive at the ‘maximum
value’ outcome.

In some cases it may be necessary to devise a scoring scheme so that the value
of each potential solution can be quantified objectively. In fact, defining the
evaluation criteria right at the start of the problem-solving process is a good idea
because it demands a clear understanding of the objectives in order to get started.

Any scoring scheme needs to be kept simple. What matters most is that each
potential solution is evaluated against the same criteria, and that these criteria are
directly related to the project objectives.

Selecting the maximum value outcome

Once the possible outcomes have been evaluated as objectively as possible,

selecting the best or maximum value outcome is relatively straightforward.
Involving others in the evaluation/selection process can often bring a different
perspective on the value of the solution. For instance, project stakeholders may
have subtly different evaluation criteria arising from different vested interests in
the outcome. The project manager will need to use judgment, but if nothing else,
the debate will provide a useful insight into how others may react to the chosen
solution.

Remember also that project stakeholders don’t always share common objectives,
particularly on large projects where there are diverse stakeholder interests.
Consequently, what may appear to be a high value solution to the project manager,
sometimes receives only lukewarm support elsewhere. In extreme cases, this
Copyright © 2016. Taylor & Francis Group. All rights reserved.

leads to unexpected consequences (for example, muted support or even outright

rejection) because insufficient attention was paid to the different stakeholder
expectations.

Validating the solution

Before we can begin to implement a solution, two levels of validation are

needed:

1. Does it achieve the objectives within the context of the framed problem?
(Does it have high intrinsic value?)

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 54 Managing Project Uncertainty

2. What are the implications of adopting the solution? (Does it also have
extrinsic value?)

The first of these requires assessment within the context of the problem (including
all the different stakeholder perspectives). But the second needs a broader view.
Are there undesirable consequences which could impact on the project objectives
in a different way? If so, the end of one problem may only be the beginning of
another.

Table 3.3 A primer for practical problem-solving

Beware of ‘self- These words often indicate variables or relationships which

evident’ and have been left unchallenged for too long. Challenging such
‘obvious’ assumptions may lead to better (or more) options than
were originally envisaged. As a rule of thumb, the more
obvious a fact, the greater the need to test its veracity.
Widen the range of Sometimes a problem is too narrowly defined and a
the problem greater range of variables needs to be considered. This
can seem counter-intuitive since we are often taught to
simplify and break problems into manageable chunks, but
more effective solutions may appear once the scope is
widened.
Non-linear causal A causal relationship is often non-linear or bounded for
relationships example, there may be a threshold above or below which
the relationship breaks down. Try to understand the nature
(and limitations) of the relationship in as much detail as
possible.
Use creative Some uncontrollable variables turn out to be within our
approaches to control when key assumptions or established wisdom
turn uncontrollable is challenged. It is worth testing the key uncontrollable
variables into variables (that is, those which have greatest influence over
Copyright © 2016. Taylor & Francis Group. All rights reserved.

controlled ones the outcome). It may open up a new range of possibilities.

Solve, compromise Some problems are intractable. No amount of effort will
or dissolve result in an acceptable solution and it is important to
realise this before squandering vast resources on it. If
no practical solution exists, consider whether there is a
compromise – a ‘good enough’ solution, particularly if
an exhaustive search for a high value outcome will be
time-consuming or costly, or acceptance criteria are too
imprecise to enable the set of outcomes to be properly
evaluated. Finally, can the problem be dissolved by
changing variables sufficiently that the problem no longer
exists?

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
 problem-solving strategies for managing uncertainty 55

Idealized Outcomes and Negative Objectives

Problems formulated from uncertainty run the risk of focusing on negative

objectives. How do we get rid of X? (where X represents some uncertainty).
Negative objectives lead to a tendency to concentrate on symptoms (the things that
we wish to avoid) and not the underlying cause, and also overlook the opportunity
to design a better solution to the problem. Business consultant Roger Ackoff
suggests that by framing problems against negative objectives ‘… we tend to walk
into the future facing the past.’ (Ackoff 1978).

This can lead to new problems arising from unforeseen consequences. In focusing
on the thing we want to get rid of, some wider implications are missed. One
example is the use of biological controls, that is, the introduction of a species to
control the threat from a particular type of pest. In some cases, this has led to a
much worse problem, that is, the control has become a virulent pest in its own
right. In 1935, cane toads were introduced into the sugar-cane producing regions
of Australia to keep down cane beetles which were eating the crops. However,
the cane toad’s own voracious eating habits, poisonous skin and ability to rapidly
colonize, left a legacy of problems which continue today, causing major ecological
damage as the cane toads own voracious eating habits.

One way of avoiding such difficulties is to refocus the problem on achieving what
can be called an idealized outcome. An idealized outcome is an aspiration, often
a restating of the project objectives. Discussions which begin with, ‘If only we
could …’ are usually focusing on idealized outcomes. They aren’t always within
reach, but sometimes just the act of turning away from negative objectives and
identifying idealized outcomes will suggest new possibilities.
Copyright © 2016. Taylor & Francis Group. All rights reserved.

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.
Copyright © 2016. Taylor & Francis Group. All rights reserved.

Cleden, David. Managing Project Uncertainty, Taylor & Francis Group, 2016. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/asulib-ebooks/detail.action?docID=438596.
Created from asulib-ebooks on 2025-02-09 21:39:54.