Had a somewhat interesting challenge presented to me recently that could make a good kata:

Given a Set of strings representing numbers and math operands, compute what the final result of a math operation would be.

Some examples

  computeFromStrings(['5','plus','5','minus','5']) // 5
  computeFromStrings(['500','times','500']) // 250000
  computeFromStrings(paramsSet) // result of operation

readability > brevity

Really this shouldn't rely on random. Or at least do random within a defined range that is known to work 100% of the time.

ah, so the test case should be rewritten really?