Chapter 2: Making Decisions — Exam

Notes

LO 2.1 — The 8-Step Decision-Making Process

Decision = making choices among two or more alternatives. Decision making is best
understood as a process, not just a single choice — even simple decisions like where to eat
lunch involve going through this process.

The 8 Steps:

1. Identify the Problem — a discrepancy between an existing and a desired condition.

Managers must be careful not to confuse problems with symptoms. A 5% drop in sales
isn't necessarily the problem — it could be a symptom of poor product quality, bad
advertising, or shifting consumer preferences. The real problem is what's causing the
symptom.
2. Identify Decision Criteria — determine what factors are relevant to resolving the
problem. Every decision maker has criteria, whether stated explicitly or not. In the
laptop example, criteria included memory/storage, display quality, battery life,
warranty, and carrying weight.
3. Allocate Weights to Criteria — criteria are rarely equally important, so assign
relative weights. A simple method: give the most important criterion a weight of 10,
then assign weights to the rest accordingly. This ensures the decision reflects actual
priorities (e.g., memory = 10, display quality = 3).
4. Develop Alternatives — list all viable options that could resolve the problem. At this
stage, creativity matters — alternatives are only listed, not evaluated yet. Amanda
identified 7 laptops: Acer Aspire E, Apple MacBook Pro, Dell XPS 13, Lenovo
ThinkPad, Lenovo Yoga, Microsoft Surface Book, Razer Blade Stealth.
5. Analyze Alternatives — evaluate each alternative using the weighted criteria from
steps 2 and 3. Multiply each alternative's score on a criterion by that criterion's
weight, then sum all weighted scores. This gives a total score for each alternative.
6. Select an Alternative — choose the alternative with the highest total weighted score.
In the example, Microsoft Surface Book scored 249 — highest among all options.
7. Implement the Alternative — put the decision into action by communicating it to
those affected and getting their commitment. Research shows that people who
participate in the decision process are more likely to support its implementation than
those who are simply told what to do.
8. Evaluate Decision Effectiveness — did the decision actually solve the problem? If
the problem persists, ask: Was the problem incorrectly defined? Were errors made in
evaluating alternatives? Was the right alternative chosen but poorly implemented?
This step may loop back to step 1.

Man City link: Their performance analysts essentially follow this process. They identified
the problem (only 1 set-piece goal in 21 matches), analyzed alternatives (studied 500+ corner
kicks from other teams), implemented new tactics (showed players video of best practices),
and evaluated effectiveness (9 goals from corners in the next 15 matches).
LO 2.2 — Five Approaches to Decision Making
1. Rationality Assumes managers make fully logical, objective decisions that maximize
value. The rational decision maker has a clear, unambiguous problem, a single well-defined
goal, knows all possible alternatives and their consequences, and always selects the option
that maximizes goal achievement. For managerial decisions, one additional assumption is
added: decisions are made in the best interests of the organization. In reality, these
assumptions are largely unrealistic — but they serve as an ideal standard.

2. Bounded Rationality More realistic than pure rationality. Managers want to be rational
but are limited (bounded) by their ability to process information. Because they can't analyze
every alternative, they satisfice — they accept solutions that are "good enough" rather than
searching for the mathematically optimal one. Example: a finance graduate accepts a decent
job offer without exhaustively searching every possible employer — the first "satisfactory"
offer ends the search. Managers still behave rationally, just within cognitive limits.

3. Intuition Decisions made on the basis of experience, feelings, and accumulated judgment.
Not irrational — it complements both rational and bounded rational approaches. Five aspects
of intuition:

 Experience-based — past experience guides the decision

 Affect-initiated — feelings/emotions guide the decision
 Cognitive-based — skills and knowledge guide the decision
 Subconscious mental processing — the subconscious mind provides input
 Values/ethics-based — ethical values or culture guide the decision

Nearly half of executives surveyed used intuition more than formal analysis. Importantly,
individuals who understood their emotions while deciding actually achieved higher decision-
making performance — so the old advice to "ignore emotions" may be wrong.

4. Evidence-Based Management (EBMgt) Systematic use of the best available evidence —

which could be hard data, expert opinions, or colleagues' experience — to improve
management practice. Essentially an attempt to operationalize rationality. Four essential
elements:

1. The decision maker's own expertise and judgment

2. External evidence evaluated by the decision maker
3. Opinions and preferences of stakeholders
4. Relevant organizational factors (context, circumstances, members)

The weight given to each element varies by situation. 81% of executives believe data should
be at the heart of all decision-making.

Man City link (Q2-14):

 Rationality: SAP Challenger Insights gives objective, data-driven analysis of

opponents' tactics — approximating rational decision making.
  Bounded rationality: Coaches can't personally track every player in every moment
during a match. Big data fills those gaps and expands the boundaries of what can be
processed.
 Intuition: Coaches and managers still apply intuitive judgment — reading the game,
sensing player morale, making split-second calls. Data informs but doesn't replace
this.
 EBMgt: Personalized post-match reports, movement analysis, tactical breakdowns,
and set-piece studies all represent systematic, evidence-based decision-making. Gavin
Fleig's team provides each player with detailed individual feedback — a direct
application of EBMgt at the individual level.

5. Crowdsourcing Relying on a network of people outside the organization's traditional set

of decision makers to solicit ideas via the internet. Useful for finding innovative solutions,
gathering customer/supplier insights, deciding what products to develop, or even who to
promote. Hershey used it to solve a chocolate shipping problem — anyone could submit a
solution; the winner got $25,000 and a collaboration opportunity.

Man City link: The 2016 global Hackathon is a direct example of crowdsourcing. The club
invited external data and football experts to create algorithms using real player data that had
never before been made available to outsiders. The winning team built a machine learning
algorithm tracking in-game decision making and received £7,000 plus collaboration with the
performance analysis team.

LO 2.3 — Types of Decisions & Decision-Making Styles

Structured vs. Unstructured Problems

Feature Structured Unstructured

Problem type Familiar, clear, straightforward New, unusual, ambiguous
Information Readily available and complete Incomplete or ambiguous
Manager level Lower-level managers Upper-level managers
Decision type Programmed Nonprogrammed
Frequency Repetitive, routine Rare, one-time
Goals Clear and specific Vague
Time frame Short Relatively long
Solution relies on Procedures, rules, policies Judgment and creativity

Programmed Decisions — repetitive decisions handled via routine approach. The "develop
alternatives" stage is skipped or minimized because solutions are already known. Three types:

 Procedure — a series of sequential steps for a well-defined problem (e.g., purchasing

procedure for ordering tablets)
 Rule — an explicit do/don't statement ensuring consistency (e.g., lateness and
absenteeism rules)
 Policy — a general guideline leaving room for interpretation (e.g., "We promote from
within, whenever possible" — the italicized word requires judgment)
Nonprogrammed Decisions — unique, nonrecurring situations requiring custom-made
solutions. Example: GE suddenly having $83 billion in offshore cash available after a tax law
change — no existing procedure covers that.

Comparing levels: Lower-level managers mostly handle programmed decisions because

they deal with routine, repetitive problems. As you move up the hierarchy, problems become
more unstructured. Upper-level managers delegate routine decisions downward so they can
focus on the unusual and difficult ones.

Man City link (Q2-13): Football managers make both types:

 Structured/programmed: Standard training schedules, travel logistics, routine lineup

decisions, media protocols — handled by procedures and rules.
 Unstructured/nonprogrammed: Mid-season tactical overhauls in response to a losing
streak, emergency transfer decisions due to injuries, adapting to a specific opponent's
unusual formation — these require creativity and judgment. Performance analysts
support especially these nonprogrammed decisions by providing data that would
otherwise be unavailable to the decision maker.

Decision-Making Styles

People differ along two dimensions when making decisions:

 Way of thinking: Rational (logical, systematic) ↔ Intuitive (creative, holistic)

 Tolerance for ambiguity: Low (need order and certainty) ↔ High (comfortable with
multiple thoughts simultaneously)

These two dimensions produce four styles:

Ambiguity
Style Thinking Characteristics
Tolerance
Fast decisions, short-term, minimal info, few
Directive Rational Low
alternatives
Careful, adaptable, handles uncertainty, best for
Analytic Rational High
new situations
Broad outlook, long-range, highly creative, many
Conceptual Intuitive High
alternatives
People-oriented, avoids conflict, relies on
Behavioral Intuitive Low
meetings, seeks acceptance

Most managers have characteristics across multiple styles. It's best to think in terms of a
dominant style plus backup styles. More flexible managers shift styles based on the situation.

Business students and managers typically score highest on analytic — because formal
education (accounting, statistics, economics, finance) emphasizes rational thinking.

LO 2.4 — Decision-Making Biases & Errors

Managers use heuristics — mental shortcuts to simplify or speed up decision making.
Heuristics are helpful for handling complexity and uncertainty, but they can systematically
distort judgment and lead to errors.

12 Common Biases:

1. Overconfidence — believing you know more than you do; holding unrealistically
positive views of yourself. The most dangerous bias. People claiming 65–70%
confidence are right only ~50% of the time. People claiming 100% confidence are
right only 70–85% of the time. Worst when dealing with issues outside your
expertise.
2. Immediate Gratification — preferring quick rewards over future payoffs. Explains
why dieting, quitting smoking, saving for retirement, and avoiding debt are all so
difficult. Immediate rewards are vivid; future costs feel distant and abstract.
3. Anchoring Effect — fixating on the first piece of information you receive as a
starting point, then failing to adequately adjust when new information arrives. First
impressions, initial price quotes, and early estimates carry disproportionate weight.
Relevant in negotiations, salary discussions, and performance evaluations.
4. Selective Perception — organizing and interpreting events based on your existing
biased perceptions. You don't see the world as it is — you see it as you are. This
affects which problems you even notice, what information you pay attention to, and
which alternatives you generate.
5. Confirmation Bias — actively seeking out information that confirms your existing
beliefs and discounting contradictory evidence. You accept confirming information at
face value but are critical and skeptical of challenging information. Very common and
very hard to overcome.
6. Framing Bias — drawing attention to certain aspects of a situation while omitting
others, thereby distorting perception. The same situation framed as "90% survival
rate" vs. "10% death rate" produces different decisions, even though the facts are
identical.
7. Availability Bias — over-weighting events that are most recent and vivid in memory.
Classic example: managers conducting annual performance reviews tend to focus on
what the employee did last month rather than the entire year.
8. Representation Bias — assessing the likelihood of an event based on how closely it
resembles another event. Drawing false analogies — seeing identical situations where
they don't exist. Can lead to stereotyping and faulty pattern recognition.
9. Randomness Bias — trying to find meaning or patterns in genuinely random events.
Decision makers struggle with accepting that some outcomes are simply beyond
prediction or control.
10. Sunk Costs Error — fixating on past investments (time, money, effort) when
evaluating current options, rather than focusing on future consequences. The sunk
cost is gone regardless of your decision — it shouldn't affect what you do next. But
people can't let it go.
11. Self-Serving Bias — taking personal credit for successes and blaming failures on
external factors. Prevents honest self-assessment and learning from mistakes.
12. Hindsight Bias — after learning an outcome, falsely believing you "knew it all
along." Makes you overestimate your predictive ability and reduces learning from
experience.

How to reduce biases:

  Increase awareness — you can't fight a bias you don't know you have
 Undergo training to recognize specific biases (research shows this works long-term)
 Pay attention to how you make decisions, not just what you decide
 Ask trusted others to identify weaknesses in your decision-making style
 Actively seek disconfirming evidence, especially when confidence is high

60% of managers believe biases they weren't aware of affected their decisions at work.

LO 2.5 — Cutting-Edge Approaches to Decision Making

Design Thinking Approaching management problems as a designer would approach design
problems. Traditional managers analyze given options and pick the highest net present value.
A design thinker asks: "What would be lovely if it existed but doesn't yet?" It goes beyond
rational analysis — incorporates emotional elements, uses observation and inquiry, and
involves collaborative, integrative problem identification. Organizations like PepsiCo have
used design thinking to fundamentally reimagine products (e.g., Pepsi Spire machine) rather
than just incrementally improving them.

Big Data Refers to huge and complex data sets that traditional software cannot process.
Enabled by cloud computing, big data allows legal research in seconds (vs. days), medical
diagnosis with greater accuracy than humans, and sports decisions informed by deep
statistical patterns. Big data has opened the door to widespread AI adoption.

Artificial Intelligence (AI) Using the power of computers to replicate human reasoning
functions. Goes well beyond simple "if-then" logic — AI can learn and solve complex
problems. Already changing consumer lives (Siri, Google Maps, Uber, self-driving cars) and
is rapidly transforming managerial decision making.

Machine Learning A branch of AI where systems learn from data, identify patterns, and
make decisions with little or no human assistance. Microsoft CFO Amy Hood used machine
learning to analyze 750,000 customers, forecast sales opportunities per product/city, and
predict which customers were at risk of leaving.

Deep Learning A subset of machine learning using artificial neural networks that simulate
the human brain — nodes connected like a web, enabling nonlinear data processing.
Example: a deep learning system trained on 100,000+ images identified melanoma with 95%
accuracy vs. 87% accuracy by 58 experienced dermatologists.

Analytics The use of mathematics, statistics, predictive modeling, and machine learning to
find meaningful patterns in data sets. The 2018 Super Bowl-winning Philadelphia Eagles
used analytics to discover that going for it on fourth downs — against conventional wisdom
— actually improved their win probability in many situations.

Currently ~24% of businesses are implementing AI; among large corporations, that number
exceeds 97%. As AI learns through experience and reflects a manager's values and goals,
decisions will increasingly approach the assumptions of rationality.
Man City — Big Data in Practice (Q2-14, Q2-15, Q2-16, Q2-17):

Q2-15 (Conditions): The performance analysis team operates mostly under risk conditions
— they have historical data and can calculate probabilities (e.g., likelihood that a certain
corner kick tactic produces a goal), but outcomes are never guaranteed. Some elements
involve uncertainty (e.g., how an opponent adapts in real time, unexpected injuries mid-
match). Pure certainty is rare in football.

Q2-16 (Quantitative data only?): No — quantitative data alone is insufficient. Numbers can
tell you how many tackles a player made but not whether they led by example in the dressing
room, adapted psychologically under pressure, or communicated effectively with teammates.
Qualitative factors — morale, leadership, creativity, team chemistry — are real and
significant. Big data should complement human judgment, not replace it. As Gavin Fleig
noted, the goal is to help players become more reflective and aware — data is a mirror, not a
verdict.

Q2-17 (Future of big data in football): Machine learning algorithms can eventually track
real-time decision making during matches, enabling live tactical adjustments. AI could
predict opponent moves before they happen based on historical patterns. Youth development
can be personalized using data about each player's developmental trajectory. The line
between performance analysis and actual coaching decisions will continue to blur — but
good judgment will always remain essential alongside any technology.

Key Definitions
Term Definition
Decision Making a choice among two or more alternatives
Problem Discrepancy between an existing and a desired condition
Decision criteria Factors that define what's important in resolving a problem
Rational decision
Logical, consistent choices that maximize value
making
Bounded rationality Rational but limited by ability to process information
Satisfice Accept solutions that are satisfactory and "good enough"
Intuitive decision
Decisions based on experience, feelings, and accumulated judgment
making
Systematic use of best available evidence to improve management
EBMgt
practice
Crowdsourcing Soliciting ideas from outside the org via internet
Structured problems Straightforward, familiar, easily defined problems
Unstructured problems New or unusual problems with ambiguous/incomplete information
Programmed decision Repetitive decision handled by a routine approach
Nonprogrammed
Unique, nonrecurring decision requiring custom solution
decision
Procedure Sequential steps used to respond to a structured problem
Rule Explicit statement of what can/cannot be done
Policy Guideline for making decisions, open to interpretation
Term Definition
Heuristics Mental shortcuts to simplify or speed up decision making
Big data Huge and complex data sets beyond traditional software capacity
Artificial intelligence Using computers to replicate human reasoning functions
AI method where systems learn from data with minimal human
Machine learning
input
Deep learning Subset of ML using neural networks to simulate the human brain
Analytics Math/stats/modeling to find meaningful patterns in data
Approaching management problems as designers approach design
Design thinking
problems

Summary — Chapter 2: Making Decisions

LO 2.1 — 8-Step Decision-Making Process

Decision = choice among alternatives; a process, not just a choice.
 1. Identify Problem — existing vs. desired condition; don't confuse with symptoms
2. Identify Criteria — relevant factors (memory, battery, weight)
3. Allocate Weights — most important = 10, rest relative
4. Develop Alternatives — list only, don't evaluate yet
5. Analyze Alternatives — score × weight, sum totals
6. Select Alternative — highest score wins
7. Implement — involve people; they'll support it more
8. Evaluate — did it solve the problem? Loop back if not

LO 2.2 — Five Approaches to Decision Making

 Rationality — fully logical, maximizes value; unrealistic in practice
 Bounded Rationality — rational but limited; leads to satisficing (good enough)
 Intuition — experience, feelings, judgment; complements rationality; 5 aspects: experience,
affect, cognitive, subconscious, values
 EBMgt — best available evidence; 4 elements: expertise, external evidence, stakeholder
opinion, org context
 Crowdsourcing — outside ideas via internet (Hershey, Man City Hackathon)

LO 2.3 — Types of Decisions & Styles

Structured vs. Unstructured: (PIMDFGT)

Problem type, Information, Manager level, Decision type, Frequency, Goals, Time frame
Feature Structured Unstructured

Problem Familiar, clear New, ambiguous

Manager
Lower Upper
level

Decision
Programmed Nonprogrammed
type

Procedure/Rule/ Judgment &

Solution
Policy creativity

Programmed types: Procedure (steps) → Rule (do/don't) → Policy (guideline, open to

interpretation)

Decision Styles — 2 dimensions: way of thinking (rational↔intuitive) × tolerance for

ambiguity (low↔high):
ABCDs

Thinkin
Style Ambiguity
g

Directive Rational Low — fast, short-term

Analytic Rational High — careful, adaptable

Conceptu High — creative, long-

Intuitive
al range

Behaviora
Intuitive Low — people-oriented
l

Business students score highest on Analytic.

LO 2.4 — Biases & Errors

Managers use heuristics (shortcuts) → useful but cause bias.

12 Biases: (OIASCFARRSSH)

1. Overconfidence — know more than you do; most dangerous

2. Immediate Gratification — quick rewards over future payoffs
3. Anchoring — fixate on first info
4. Selective Perception — biased interpretation
5. Confirmation — seek info that agrees with you
6. Framing — highlight some aspects, omit others
7. Availability — over-weight recent/vivid events
8. Representation — false analogies
9. Randomness — finding patterns in random events
10. Sunk Costs — fixate on past investments
11. Self-Serving — credit successes, blame failures externally
12. Hindsight — "I knew it all along"

Reduce by: awareness, training, reflect on how you decide, seek disconfirming evidence.

LO 2.5 — Cutting-Edge Approaches

  Design Thinking — think like a designer; collaborative, emotional + rational; asks "what
doesn't exist yet?"
 Big Data — huge complex data sets; beyond traditional software
 AI — computers replicating human reasoning; learns and solves complex problems
 Machine Learning — learns from data, minimal human input
 Deep Learning — neural networks simulating human brain; subset of ML
 Analytics — math/stats/modeling to find patterns

Man City Case — Key Points

Question Answer

Q2-13 Decision Both structured (training schedules) and unstructured

types (tactical overhauls, transfers)

Rationality=SAP data; Bounded=fills gaps coaches

Q2-14 Big data
can't; Intuition=coaches still judge;
+ approaches
EBMgt=personalized reports

Q2-15 Mostly risk (probabilities known); some uncertainty

Conditions (opponent behavior)

Q2-16 Quant No — qualitative factors (morale, leadership) matter too;

only? data complements, not replaces judgment

Real-time tracking, predict opponent moves,

Q2-17 Future
personalized youth development