The top 5 things I’ve learned about Testing from Mythbusters
(note: this article has a few spoilers)
1. Performance and Stress testing techniques
Much of what Adam and Jamie bust are related to performance and stress issues. For example, in episode #10 they attempt to see at what point the real axle of a police car can be ripped out the back. In this episode, we first hear what their ongoing strategy is: “replicate the circumstances, then duplicate the results”.
Applying this to Perf and Stress Testing (and I’m simplifying things here greatly), the first step is to replicate the typical environment(s). The next step, is to keep pushing the limits until the system exceeds the limits placed on it, e.g. ripping the axle out the back of a police car. The first step gives you the information most valuable for something usable, while the second step provides further information about the boundaries of the environment something can operate in.
2. Create different small prototypes first
Whenever the Mythbusters tackle a problem, they tend to follow some methods that reduce the problem down to a small scale, or into many smaller problems to solve. We see this demonstrated, for example, in the lead balloon episode where Adam and Jamie first make tiny small scale mockups in order to try out different theories. All the different theories they try demonstrate a corollary to this method: always brainstorm many different theories and then try out the best ones. With this method, the Mythbusters save themselves significant time and money in the long run because the small scale experiments are quick, easy, and cheap to execute. Also, in the case of film myths, they consistently break the myths into sub-myths that can be individually tested, otherwise they would often be replicating a Rube Goldberg chain reaction doomed for failure.
3. Refine testing through an iterative model
This method proceeds from the previous two methods. Once the Mythbusters have a small prototype that seems compelling, they begin to expand the scale of testing. Consider if they always started big. They might spend a significant amount of time and resources and ultimately fail–and worse, they would be back where they started. Conversely, repeating and expanding experimentation yields a constant stream of information that the Mythbusters feed back into the next iteration. Even if they sort of fail, they can just refine the previous iteration and ultimately get to the end of the episode with a large body of data. Often, as in the concrete glider episode, they have learned a great deal about the various materials involved, they have generated new methods and tools, and they have recognized various patterns within the data. At the end of the episode, their gliders crashed and crumbled, but Adam and Jamie called the myth “plausible” because they had done enough testing to discover working methodologies and to see the promising affects of the force of lift on the gliders (among other things). Had they created a full scale concrete glider or lead balloon at the very beginning, they certainly would have failed before completing any end to end working model and before getting any useful data. That likelihood of failure is, in fact, the reason for the myth, and it was by creating steady refinements that Adam and Jamie eventually created a full scale floating lead balloon.
4. Use design patterns
This is a meta-method in testing, which the Mythbusters demonstrate whenever they need something to go flying through the air at a specific speed. They have learned that the way to send something through the air is with the “air cannon design pattern”. In other words, use an air cannon to shoot the thing through the air, because it’s easy to control the air compression, easy to build, easy to maintain, and easy to re-use. The specific materials don’t matter. They can apply the Air Cannon Design Pattern to any situation, building any size they need, shooting at any specific speed (more or less), using any materials–to create an instance of an Air Cannon. This is the essence of design patterns, and the Mythbusters execute it with elegance.
5. Reuse your tools
One of the benefits the Mythbusters have taken advantage of over time, by using design patterns, is the ability for them to re-use their tools in all sorts of situations. Over and over, we see them re-use various air cannons. Over and over, we see them bring Buster the test dummy into episodes so that they can shock him, shoot him, drop him, or explode him. They certainly got their money’s worth with Buster, and the more they use him, the more they knew how to use him again. ~