Pretty much an exact duplicate of at least the following already approved Kata, since the exact solution (modulo the function name) can be re-used between each Kata:

• https://www.codewars.com/kata/57a049e253ba33ac5e000212
• https://www.codewars.com/kata/54ff0d1f355cfd20e60001fc
• https://www.codewars.com/kata/5694cb0ec554589633000036

I've unpublished the Kata for now. Please refer to Creating your first Kata for a comprehensive guide on authoring high-quality Kata.

Ok, you are almost there, but the problem is that you are still calling the user solution in the same unshuffled order. You should move the dogs_race(i) call to the assertion: test.assert_equals(dogs_race(case[0]), case[1])

I haven't thought about that. I have added shuffling to my test cases.

Yeah, random tests are much less important for easy kata, however it is still good authoring practice, and almost always required before a kata has a chance of approval.

You can simply use your current approach to build a list of input-answer pairs, then use random.shuffle to make the order random, and run through those

good point, I have changed the description and removed unnecessary import

Random tests are not random

Okay, so you do not want this solution to be submitted.

I think you have your assertion the wrong way round.

But, looking at the solutions, I think you have a bigger problem. You will have to force solvers to do what you want, and not just ask them nicely. Which is a problem, because kata generally are about the result, and not the way you get there.