Harold in Trouble

Case Study • Story • Outcome • Creation Process • Downloads

Case Study

Authors: Georg Struck, Steve Hoffmann, Ulrike Spierling, Kim Dung Dang, Ronan Champagnat

The “Harold in Trouble” story was meant as an educational approach to teach authors the concepts of planning. Because of that its first implementation was as a physical card game so that authors could experience planning from the side of the planning algorithm itself by recreating its working steps by hand. Later it was also implemented into the EmoEmma authoring tool and also into a Linear Logic representation for further exploration of authoring concepts.

Story

The story is about a clumsy wannabe womanizer and secret agent assistant named Harold who has to stay calm during a garden party of a super villain meanwhile his boss, the top agent Number Seven, tries to negotiate with the evil Silvertoe. Trying to seduce a woman he creates chain reactions of chaos what makes Silvertoe angrier and angrier.

Outcome

The figure below shows an example layout of a play session on the left and a final card for the physical card game on the right. During the design phase it is simple to use a piece of paper and write down actions, preconditions and effects. Because this is far from a software implementation it is also possible to leave some of the fields open to fill them later in the process but still be able to play around and get an impression of the outcome.

The picture below gives an impression of a play session with the card game at one of several workshops.

Creation Process

The table below shows an early design stage in which alternative actions were explored by using the creative principle of abstraction. During a brain storming phase different possibilities to “seduce women” from the point of a “womanizer” are explored. Each of them is placed on different abstraction level, making it possible to regroup them to find out if some actions are super or subgroups of others. This allows later filling gaps and finding in these way alternatives that were not thought about. It also helps in further design steps to model to model the story by beginning with more abstract and because of that more simple actions to prevent a too early explosion of options. If a working abstract story structure is found it later can be refined step by step.

To create chain reactions multiple chunks of actions have to be found and connected with each other, keeping in mind that they should be interchangeable as much as possible. The figure below on the left shows one draft of such chain reactions. The figure on the right shows a more concrete chain, taken from the left figure and fleshed out with some details.