Model: naming tests
Naming tests
- snake case is more readable than camel case (when > 30 characters, which tests usually are)
- "should" is like training wheels
name tests as propositions

name test as Method_input_result

"All models are wrong, but some are useful"