BDD & The Feynman Technique

BDD a team-level Feynman Technique for shared learning

By Antony Marcano

Nobel Prize winner and accomplished physicist Richard Feynman, was renowned for his ability to explain complex subjects.

It was for this reason that he was known by many as “The Great Explainer”.

“If you can’t explain it simply,

you don’t understand it enough.” *

Feynman learned by trying to explain things. If he couldn’t explain it simply, he didn’t understand it enough. It is from this that the Feynman Technique for learning emerged:

1. Pick a concept — write it at the top of a fresh page in your notebook
2. Pretend to teach it to a student — explain it on paper as if you are teaching it
3. Study it more when you get stuck — keep studying until you can resume explaining it
4. Simplify and use analogies — relate it to things people already understand

This is what we are doing at a team level when we’re practicing BDD**…

If we can’t tell the user story simply,

we don’t understand it enough.

The purpose of writing a customer scenario isn’t to produce an artefact or an automated test. The purpose of writing a customer scenario is to drive a conversation, drawing it down to the right level of detail to establish a shared, coherent understanding — to learn together.

This illustrates how we see and practice BDD.

If the scenarios that tell variations on a user story can’t be explained simply, we don’t understand the story enough as a team.

Each edit, each refinement we try to make, takes the conversation and our understanding deeper. We gradually get closer to a better understanding of what the customer is trying to achieve and why.

When we can articulate our understanding of a scenario simply and our customer agrees that we’re aligned, that’s a good sign that we understand it enough to take the next step — implementing just enough to see that scenario pass. Now we’re stuck and it’s time to explore the next scenario.

The simpler the scenario, the more the team understands what the customer is trying to do.

An example of how we might describe configuring Slack’s automated Slackbot responses.

The process of reflecting and evolving our understanding in the actual product continues into the process of implementation. Unit-level specs (unit tests) drive the conversation with our pair-programming partner (or mob) — capturing our understanding of how we’ll achieve the outcome we’re looking for — and expressing the way we’ve achieved it through our system metaphor**.

This process is iterative. There will still be gaps in our understanding — we don’t need it to be perfect first time, every time… Addressing that is exactly what short cycles and frequent feedback are for.

TL;DR

BDD — Feynman Technique analogy…

1. Pick a concept — take the next user story;
2. Pretend to teach it to a student — outline the example scenarios, elaborate one, discuss, refine, edit until it’s simply articulated in the scenario and eventually in your product code;
3. Study it more when you get stuck — iterate, return to the whiteboard together, including the customer/product owner, discuss just enough to resume explaining it in a scenario, your unit-specs and product-code;
4. Simplify and use analogies — Refine the story a user would like to tell through the variations in each scenario, refactor the code to simply express its intent by analogy, through a system-metaphor.

Footnotes

*This quote is often attributed to Albert Einstein and sometimes to Richard Feynman, however the origin of the quote is disputed by many, hence we’ve not given explicit attribution.

** Arguably, XP as a whole — including system metaphor — is a software development approach analogous to the Feynman Technique .

Explore ways you can unlock similar insights by contacting us via RiverGlide.com