ArgonDigital - enterprise automation experts

Share This Post

Requirements Models: Decision Trees and Decision Tables, a Choose Your Own Adventure™ Blog! Part 2 of 2

Welcome back!
If you chose to read about Decision Tables in Part 1, go to paragraph 2 below.
If you are continuing your journey from Decision Trees, go to paragraph 3 below.

2. Ah, you chose Decision Tables, a more compact method to view all of the available decisions at once. I created the Decision Table after the Decision Tree was completed. Strictly speaking, since all of the decisions in the book are ordered (in that you have to make the first decision before you can move on) a Decision Table shouldn’t be very useful for this book.

However, I found it extremely interesting to see all of the decisions and possibilities lined up in a tabular format. For one thing, the Decision Table showed me right away that, though there only 32 different endings, there were 44 ways to get to one. The full Decision Table is shown below; for readability, I separated the table into two (please click on each Decision Table to see the full size version.)

Part 1:

Part 2:

This decision table reveals several interesting things (interesting to me, at least.)

Of the 32 different endings, 14 had the character back home with his parents, 7 had the character stuck in a parallel world in a negative way, 6 had him stuck in a parallel world in a positive way, and 5 had him dying in another world. These were statistics that would have been difficult to determine from the Decision Tree but easy from the Decision Table.

For the total possible 44 ways to get to an ending, for the respective categories above, the numbers were 17, 11, 10, and 6.

With the table, it was also easier to choose an ending (say stuck in a parallel world in a positive way), choose a column and have all the decisions I would need to make to get there, starting with the first decision at the top. Additionally, by counting the number of decisions in a column, I could choose the longest or shortest path to get me to my desired outcome.

However, the Decision Table, by itself, is harder to read and follow than the Decision Tree; ensuring completion was more difficult since there are so many exclusionary decisions. If the reader made every decision, however, ensuring completeness would be easier with the Decision Table; you would need 2^n columns where n is the number of decisions.

Some of the major advantages and disadvantages of Decision Tables are listed below.

Advantages

  • Statistics: For counting the number of different endings, to determining what endings were more likely, the Decision Table makes this easy since Microsoft Excel will do most of the calculations for you.
  • Backwards traceability: The Decision Table presents the decisions in such a way that choosing an ending and finding what decisions need to be made is easy.
Disadvantages
  • Ensuring completion: Generally speaking, Decision Tables are better than Decision Trees at ensuring completion and that you have analyzed every possible combination of decisions. However, in this case, because there were so many decisions that only applied if other decisions were made, the only way to ensure completeness of this Decision Table is to compare it to the Decision Tree.
  • Size and Aesthetics: Because every possible series of decisions and outcomes is a column, the Decision Table becomes too wide to be easily inserted into Microsoft Word or Powerpoint documents. Also, with everything in tabular form, the Decision Table looks very dense and is more difficult to digest.
Continue on to paragraph 3.

3.  So, this series of blog posts weren’t’t very much of a “Choose Your Own Adventure™,” with just one decision to be made, but I think “The Antimatter Formula” by Jay Leibold showed some of the advantages and disadvantages of Decision Trees and Decision Tables quite well.

For this example, a Decision Tree is probably best, since it is easier to read and follow. However, if you wanted to run some statistical analysis of the various endings, a Decision Table is incredibly useful.

In the end, I’ve found that using Decision Tree and Decision Tables in tandem, where possible, leads to the best results in both readability and completeness. Doing this experiment also made me wonder if Decision Trees and Decision Tables, or something like them, weren’t used by the authors when creating the story…

For more information, please read our other blog posts about Decision Trees and related visual models!

Requirements Models: Decision Trees and Decision Tables, a Choose Your Own Adventure™ Blog! Part 2 of 2

More To Explore

AI to Write Requirements

How We Use AI to Write Requirements

At ArgonDigital, we’ve been writing requirements for 22 years. I’ve watched our teams waste hours translating notes into requirements. Now, we’ve cut the nonsense with AI. Our teams can spend

ArgonDigital | Making Technology a Strategic Advantage