Uncertainty and Chance

A good quote from the book of Smart Choices, "The quality of our decision making is not judged by the quality of consequences."

The authors put it, you may end up:

• A smart choice, a bad consequence: the decision is fine, but it is only things happen. The rare bad case scenario happens which ruins the result. How do you know the rare bad thing could happen before make the decision? We never know! This is uncertainty.
• A poor choice, a good consequence: even the decision is poor but outcome is good, it is just because of luck. 

Once you have made the decision right, the outcome may be subject to uncertainties.

 

 

What uncertainty means to us?

The authors put it: "you may know what might happen, but you won't know what will happen." This is because of uncertainty. 

Uncertainties can be arisen from: 

• (a) information/ data uncertainties, 
• (b) model uncertainties, 
• (c) calculation/ mathematical uncertainties, 
• (d) result/ outcome uncertainties, etc. 

Uncertainties complicate decision making process by introducing an additional layer (or multi-layers) of concerns, especially for linked decisions. In an independent single decision making process, it may involve different uncertainties with varying levels of importance. We can't make uncertainties go away, but we can address them. 

 

How to Address Uncertainties?

In any decision making process, address the following three main points:

• Key Uncertainties: Be able to identify the key uncertainties: 
• list all the uncertainties out which are significantly influence the consequences of the alternative and ignore the rest. 
• Think what will happen to the outcome if the uncertainty varies at different degrees. 
• If it substantially affects the extent of the outcome, that is the key uncertainty you must address, and you can't ignore it. 
• Otherwise, it is not significant, you may ignore it.
  
• Possible Outcomes: Identify the possible outcomes from the identified key uncertainties.
• Assign Chances: Judge the chance of happening the possible outcomes. This can be done by:
• your personal judgement (which inevitably will subject to heuristics and human biases), 
• collect enough available existing information, 
• collect new data (you may explore what it is like by trial period, or by actually experiencing it),
• ask experts (before asking, you better have your own thinking or judgement first), 
• break uncertainties into components and assign a chance to each components.

 

Example 3:

Your company wants to bid a big project from an corporation, which would produce a high profit but it also requires a high cost to invest in for the preparation of the bidding proposal/ research and else. The corporation's response becomes the key uncertainty for the success of bidding the project. 

• Decision problem: Whether to bid the project or not?
  
• Key uncertainty: The corporation's response  
• Possible outcomes: Get a full contract / partial contract / no contract
• Assign chances: This is the most difficult task in the process. By personal judgement, you know that with the capability of the company this is least likely that your company will have no contract. For full contract, the corporation may think your company does not afford to hold the project alone, so also not likely to have a full contract. For partially grant a contract, you judge that it is most likely would be the case.
  

Now, a simple table in this case can be drawn (Figure 1). You can see the clear good and bad for partial and no contracts, but a little vague for full contract. What do I mean by "not likely"? For this problem, I will articulate it in another article <Subjective Probability Estimation>.

Figure 1. Risk profile using descriptive chance and consequence. 

 

Incorporating Expected Value Theory for Decision Making

I incorporate the Expected Value theory for quantifying consequence, but it seems possible to me to use it for further clarifying the consequences involved in the uncertainty. So, I have made another risk profile for comparison, using Expected Value Theory, by converting all qualitative descriptions to quantitative measures, you can see the comparison is much clearer now (Figure 2).  

Reference: Not bidding ($0)  ← No cost put in

• Full Contract: Expected Value = $168,000 (gains) - $20,000 (cost) = $148,000
• Partial Contract: Expected Value = $412,500 (gains) - $20,000 (cost) = $392,500
• No Contract: Expected Value = $0 (gains) - $20,000 (cost) = - $20,000

Comparing the reference point of not bidding the project ($0), the outcome of either having full contract ($148,000) or partial contract ($392,500) are better than not bidding. The outcome of having no contract (-$20,000) is worse than no bidding. In view of chance, there is only 10% chance of having no contract, that means 90% chance would turn out be either partial or full contract. Either one is better than not bid. 

In Expected Value theory, we should seek the choice with maximum Expected Value. So, the decision is clear. You should go for bidding the project. The advantage of adopting Expected Value for your decision making process is that it removes the ambiguous terms and the arbitrary consequence descriptions, making a decision simple and clear.

Figure 2. Risk profile using Expected Value theory with quantitative measures.

 

Conclusion

To me, identifying the key uncertainties and assigning chances for each of them are the most difficult parts of the decision making process because it involves subjective judgement. The 'uncertainties' mentioned here are only arisen from the outcome uncertainty. 

The other difficulty is that how to translate the emotional rewards/penalties to a quantifiable value (usually money). As in Example 3, I adopted the concept of willing-to-pay (WTP) and willing-to-accept (WTA) to help me do the job. This is the concept of trade-off.

I have also applied the Expected Value Theory to help making a more systematic decision.

When the above elements are involved in your decision making process, there are always errors produced, however accurate, precise and sophisticated the model you used, so it is wise to always budget for errors - always!

More Topics:

Smart Choices (1) | Focus and Defining Problem

Smart Choices (2) | Eight Elements in PrOACT Approach

Smart Choices (3) | Comparing with Consequence Table

Trade-offs in Decision Making

Decision Tree

Subjective Probability Estimation 

Assessing Risk Tolerance for Decision Making

風險決策的兩個理論: 期望值 & 期望效用

Reference

John S. Hammond, Ralph L. Keeney, Howard Raiffa, Smart Choices - A Practical Guide to Making Better Life Decisions, Broadway Books, 1999.

 

==================================

Disclaimer

No part of the post may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without written permission from the blog creator, Snow Hill 高山雪.

This is a personal blog, the blog creator does not warrant, guarantee, or assume responsibility for the accuracy or reliability of any information posted. Under no circumstances shall the blog creator has any liability to you for any loss or damage of any kind incurred as a result of the use of, or reliance on, the information of the site or blog posts. The use of the site and your reliance on any information on the site is solely at your own risk.  

 © Copyright 2020-2022 高山雪 Snow Hill. All rights reserved.