Skip to content

Testing Anti-Patterns

"A test suite is a body of code, and code rots. The difference is that when production code rots you get bugs; when test code rots you stop trusting the tests — and then you stop running them."

The anti-patterns in this chapter live in the tests, not the production code. They are the named shapes of bad tests — drawn mostly from Gerard Meszaros's xUnit Test Patterns and the test-smell literature that followed.

Every one of them makes the suite lie: it stays green while the code is broken, flips red when nothing changed, or grows so slow and brittle that the team quietly stops running it. A suite you don't trust is worse than no suite — it costs maintenance and gives false confidence.

The Six Anti-Patterns

Anti-pattern Symptom Primary cure
Fragile Tests A test breaks on a change that didn't change behavior Test behavior through the public API, not internals
Flaky Tests Same code, same test, different result run to run Control time, await conditions (not sleep), isolate state, seed randomness
Mystery Guest A test depends on data you can't see in the test Make fixtures local and explicit; build state in the test
Assertion Roulette Many unlabelled assertions; on failure you can't tell which One logical assertion per test, or descriptive messages
Slow Tests The suite takes so long nobody runs it before pushing Test pyramid, fakes over real I/O, parallelize, slice
Over-Mocking Tests assert on mocks, not behavior; refactors break them Mock only at real boundaries; prefer fakes/real objects

Why These Sit in the Anti-Patterns Roadmap

Test smells are anti-patterns by the strict definition: a recurring "solution" (a way of writing the test) that looks reasonable, even diligent, but reliably produces a worse outcome — a suite that obstructs change instead of enabling it. They pair with the same cures as the rest of this roadmap: a code smell to name, and a better structure to move toward.

graph TD OM[Over-Mocking] --> FR[Fragile Tests] MG[Mystery Guest] --> FL[Flaky Tests] SL[Slow Tests] --> SKIP[Suite gets skipped] FR --> SKIP FL --> SKIP SKIP[Nobody runs it → no safety net → bugs ship]

A fragile or flaky suite gets ignored; an ignored suite is not a safety net; without a safety net, refactoring stops and the production anti-patterns in the rest of this roadmap grow unchecked. Good tests are what make the other cures affordable.

File Suite

Each anti-pattern ships the standard 8-file suite:

File Focus
junior.md What it looks like, why it's bad
middle.md When it creeps in, what to do instead
senior.md Fixing it across a real suite
professional.md Edge cases, trade-offs, when it's tolerable
interview.md 30+ Q&A
tasks.md Exercises with solutions
find-bug.md Spot the test smell
optimize.md Refactor the bad test

Code examples are in Go, Java, and Python, matching the rest of Code Craft.

  • Anti-Patterns (parent) — the production-code anti-patterns these tests are supposed to catch.
  • Refactoring → Code Smells — test smells are code smells in test code.
  • The unit-testing-patterns, mocking-strategies, test-data-management, and integration-testing skills cover the positive patterns these anti-patterns violate.