individual assignment for Kesha grey only
© 2018 Boston University Metropolitan College Slide 1
Tutorial for Preparation of Assignment 2 and for Building a Decision Tree in Microsoft Excel Add-In ‘TreePlan’
Table of Content
Task 2-1: Payoff Table Preparation & Simplified Decision Tree
Task 2-2: EMV Calculations
Task 2-3: Sensitivity Analysis
Task 2-4: Decision Tree in Excel Add-In ‘TreePlan’
BU MET AD 715 Quantitative & Qualitative Decision-Making
Example of Decision Tree Analysis
© 2018 Boston University Metropolitan College Slide 2
Problem Description:
A business owner is considering whether to open a new shop in City A.
There are three decision alternatives:
(1) Open a small shop
(2) Open a medium-sized shop
(3) Do not open a shop
The amount of payoffs (profit) depends on the following market conditions:
(1) Good Market
(2) Average Market
(3) Bad Market
The probabilities for Good Market, Average Market, and Bad Market are predefined.
For the purpose of the sensitivity analysis, the probability of Bad Market condition is given.
Payoff Table: Template
© 2018 Boston University Metropolitan College Slide 3
Task 2-1 Payoff Table Preparation & Simplified Decision Tree
The estimated payoffs and probabilities are shown as follows:
Alternatives States of Nature
Good Market Average Market Bad Market
Small Shop $40,000 $20,000 -$15,000
Medium-sized shop $70,000 $30,000 -$50,000
No shop $0 $0 $0
Probability 0.3 0.4 0.3
Probability for Sensitivity Analysis
P 0.6 - P i = 0.4
0.6
Payoff Table: Template
© 2018 Boston University Metropolitan College Slide 4
Task 2-1 Payoff Table Preparation & Simplified Decision Tree
The estimated payoffs and probabilities are shown as follows:
Alternatives
States of Nature
Good Market Average Market Bad Market
Small Shop $40,000 $20,000 -$15,000
Medium-sized shop $70,000 $30,000 -$50,000
No shop $0 $0 $0
Probability 0.3 0.4 0.3
Task 2-1: Payoff Table Solution
© 2018 Boston University Metropolitan College Slide 5
Task 2-1 Payoff Table Preparation & Simplified Decision Tree (continued)
Based on the payoff table, we can draw the simplified decision tree (without
probabilities and EMVs), where 1 is a decision node, and 2, 3, and 4 are chance nodes:
Task 2-2: EMV Calculation
© 2018 Boston University Metropolitan College Slide 6
Task 2-2: EMV Calculation
Based on the payoff table, the EMV for each decision alternative could be calculated as
below:
(1) EMV (Small Shop) =
($40,000)*0.3 + ($20,000)*0.4+(-$15,000)*0.3 =$15,500
(2) EMV (Medium-Sized Shop) =
($70,000)*0.3+($30,000)*0.4+(-$50,000)*0.3 =$18,000
(3) EMV (No Shop) =
$0*0.3+$0*0.4+$0*0.3 =$0
Task 2-2: Solution – EMV Calculation
© 2018 Boston University Metropolitan College Slide 7
Task 2-2: EMV Calculation (continued)
Payoff Table Results:
Alternatives
States of Nature
EMV Good Market Average Market Bad Market
Small Shop $40,000 $20,000 -$15,000 $15,500
Medium-sized shop $70,000 $30,000 -$50,000 $18,000
No shop $0 $0 $0 $0
Probability 0.3 0.4 0.3
Task 2-2: Solution
© 2018 Boston University Metropolitan College Slide 8
Task 2-2 Solution: The best EMV is EMV (Medium-sized shop)
Alternatives
States of Nature
EMV Node Good
Market Average Market
Bad Market
Small Shop $40,000 $20,000 -$15,000 $15,500 2
Medium-sized shop
$70,000 $30,000 -$50,000 $18,000 3
No shop $0 $0 $0 $0 4
Probability 0.3 0.4 0.3
Final Decision EMV Node
Medium-sized Shop $18,000 1
Task 2-3: Sensitivity Analysis
© 2018 Boston University Metropolitan College Slide 9
Task 2-3: Sensitivity Analysis:
Let’s assume that the probability of Bad Market condition is i = 0.4
Use sensitivity analysis to determine the probability range for Good Market which would
change the business owner's decision. Draw a sensitivity chart and find the probability
for the cross points.
Task 2-3: Sensitivity Analysis
© 2018 Boston University Metropolitan College Slide 10
The given information is that the probability of Bad Market is 0.4. Therefore, the total
probability of Good Market and Average Market is 0.6. If we set P as the probability of
Good Market, the probability of Average Market would be (0.6-P).
Alternatives
States of Nature
Good Market Average Market Bad Market
Small Shop $40,000 $20,000 -$15,000
Medium-sized shop $70,000 $30,000 -$50,000
No shop $0 $0 $0
Probability P 0.6 - P 0.4
Task 2-3: Sensitivity Analysis
© 2018 Boston University Metropolitan College Slide 11
EMV (No shop) = $0*P+$0*(0.6-P)+$0*0.4 = $0
EMV (Medium-sized shop)= ($70,000)*P+($30,000)*(0.6-P)+(-$50,000)*0.4
= $40,000*P-$2,000
EMV(Small Shop) = ($40,000)*P + ($20,000)*(0.6-P)+(-$15,000)*0.4
= $20,000*P+$6,000
Alternatives
States of Nature EMV
Good Market Average Market Bad Market
Small Shop $40,000 $20,000 -$15,000 $20,000*P+$6,000
Medium-sized shop
$70,000 $30,000 -$50,000 $40,000*P-$2,000
No shop $0 $0 $0 $0
Probability P 0.6-P 0.4
In summary, we have the equations for the three EMVs (in dollars), where P is the probability of Good Market:
EMV (Small Shop)=$20,000*P + $6,000
EMV (Medium-sized shop)=$40,000*P - $2,000
EMV (No shop)=0
When P = 0: EMV(Small Shop)=$6,000
EMV(Medium-sized Shop)=-$2,000
EMV(No shop)=$0
When P = 0.6: EMV(Small Shop)=$18,000
EMV(Medium-sized Shop)=$22,000
EMV(No shop)=$0
Now, we can use these points to draw the lines for EMVs to do sensitivity analysis.
Task 2-3: Sensitivity Analysis
© 2018 Boston University Metropolitan College Slide 12
Task 2-3: Sensitivity Analysis
© 2018 Boston University Metropolitan College Slide 13
Based on the previous equations, here are the three lines representing the EMVs for each
decision alternatives respectively:
Cross Point 2: P = 0.4
Cross Point 1: P = 0.05
EMV (Small Shop)
EMV (Medium-sized Shop)
EMV (No Shop)
($5,000)
$0
$5,000
$10,000
$15,000
$20,000
$25,000
0 0.1 0.2 0.3 0.4 0.5 0.6
E M V
P (Good Market)
Sensitivity Analysis
$40,000*P-2000 = 0
P = 2,000/$40,000 = 0.05
$40,000*P - $2,000 = =$20,000*P + $6,000
$40,000*P - $20,000*P = = $6,000 + $2,000
$20,000*P = $8,000
P = $8,000/$20,000 = 0.4
Task 2-3: Solution
© 2018 Boston University Metropolitan College Slide 14
From the sensitivity chart, it’s obvious to see the cross points:
When P = 0.05 EMV (Medium-sized Shop) = EMV (No Shop) = $0;
When P = 0.4 EMV (Small Shop) = EMV (Medium-sized Shop) = $14,000
Probability ranges for Good Market:
When 0 < P(Good Market) < 0.4 EMV (Small Shop) is the highest EMV, opening a
small shop is the best choice.
When 0.4 < P (Good Market) < 0.6 EMV (Medium-sized Shop) is the highest EMV,
therefore opening a medium-sized shop is the
best choice.
Task 2-4: Decision Tree in Excel Add-In ‘TreePlan’
© 2018 Boston University Metropolitan College Slide 15
Use the following Tutorial for Building a Decision Tree in Excel Add-in ‘TreePlan’:
A Step-By-Step Approach
Step 1: Getting Started
Step 2: Adding Branch
Step 3: Naming Alternatives
Step 4: Adding chance node
Step 5: Naming Alternatives
Step 6: Copy Subtree
Step 7: Paste Subtree
Step 8: Inserting Values
© 2018 Boston University Metropolitan College Slide 16
What is ‘TreePlan’?
TreePlan = The Decision Tree Microsoft Excel Add-In
‘TreePlan’ helps you build a decision tree diagram in an Excel worksheet.
How to Get Access to Decision Tree using Excel Add-in ‘TreePlan’?
Open V-Lab: http://www.bu.edu/metit/hw-and-sw/virtual-labs/ and log in to your BU account
Click excel 2016 and open an new excel worksheet.
Choose Add-Ins Decision Tree.
Tutorial for Building a Decision Tree in Microsoft Excel Add-In ‘TreePlan’
© 2018 Boston University Metropolitan College Slide 17
Step 1: Getting Started. In a new worksheet, select cell A1. From Add-Ins tab choose
Decision Tree from the Menu Commands group, then click New Tree
Tutorial for Building a Decision Tree in Microsoft Excel Add-In ‘TreePlan’
© 2018 Boston University Metropolitan College Slide 18
Step 2: Adding Branch. Now the decision tree has two alternatives. Click cell B5,
from Add-Ins tab choose Decision Tree from the Menu Commands group, click
Add Branch and click OK.
Tutorial for Building a Decision Tree in Microsoft Excel Add-In ‘TreePlan’
© 2018 Boston University Metropolitan College Slide 19
Step 3: Naming Alternatives. Entering “Open a small shop” in cell D2, “Open a
medium-sized shop” in cell D7, and “No shop” in cell D12 instead of ‘Alternative
1’, ‘Alternative 2’ and ‘Alternative 3’.
Tutorial for Building a Decision Tree in Microsoft Excel Add-In ‘TreePlan’
© 2018 Boston University Metropolitan College Slide 20
Step 4: Adding chance node. Click cell F3, choose Decision Tree from the Menu
Commands group. Then select Change to event node. Select Three in the
Branches section and Click Ok.
Tutorial for Building a Decision Tree in Microsoft Excel Add-In ‘TreePlan’
© 2018 Boston University Metropolitan College Slide 21
Step 5: Naming Alternatives. Entering “Good Market” in cell H2, “Average
Market” in cell H6, and “Bad Market” in cell H11 instead of ‘Outcome 1’,
‘Outcome 2’ and ‘Outcome 3’.
Tutorial for Building a Decision Tree in Microsoft Excel Add-In ‘TreePlan’
© 2018 Boston University Metropolitan College Slide 22
Step 6: Copy Subtree. Click on cell F8, choose Decision Tree from the Menu
Commands group. Select Copy Subtree.
Tutorial for Building a Decision Tree in Microsoft Excel Add-In ‘TreePlan’
© 2018 Boston University Metropolitan College Slide 23
Step 7: Paste Subtree. Click on cell F18, choose Decision Tree from the Menu
Commands group, then select Paste subtree.
Tutorial for Building a Decision Tree in Microsoft Excel Add-In ‘TreePlan’
© 2018 Boston University Metropolitan College Slide 24
Step 7 (Continued): Click on cell F8, Copy Subtree, and click cell F33 and Paste the
Subtree, the decision tree will be shown as the following:
Tutorial for Building a Decision Tree in Microsoft Excel Add-In ‘TreePlan’
© 2018 Boston University Metropolitan College Slide 25
Step 8: Inserting Values. List the values for each branch based on the given payoff
tables.
o In cell H1, H11, H16, H26, H31 and cell H41, input 0.3;
o In cell H6, H21 and cell H36, input 0.4.
o In cell K3, K8 and cell K13, input value $40,000, $20,000, -$15,000 respectively.
o In cell K18, cell K23, and cell K28, input value $70,000, $30,000, -$50,000
respectively.
The EMV of the decision tree will be calculated automatically by Excel.
The best EMV is $18,000, which is opening a medium-sized shop.
Tutorial for Building a Decision Tree in Microsoft Excel Add-In ‘TreePlan’
© 2018 Boston University Metropolitan College Slide 26
Step 8 (Continued): EMV Calculation by Excel
• Excel Add-in ‘TreePlan’ will automatically calculate the EMV for each chance node and give the best EMV of the each decision node.
• The best EMV is $18,000. Therefore, open a medium-sized shop is the best choice.
Tutorial for Building a Decision Tree in Microsoft Excel Add-In ‘TreePlan’