Random test cases added.

Noting some solutions are invalid, without saying what solutions you are talking about, is like saying "this kata has a problem", it may be true, but without stating the issue exactly, it isn't an issue as it can't be resolved.

.

seems resolved, now.

Maybe you agree with it, but I do not. For some reason the author mentioned (though didn't test) that ending up in a tie is neither a win nor a lose. Having N points and rolling 6 when the opponent has N+5 and rolls 1, you end up in a N + 6 = N+5 + 1 situation. The only way to truly resolve this issue is either never generate such situations or rewrite the kata so the users should handle opponent - player == 5 situation separately.

Agreed. I updated the 3 versions (...and invalidated all the previous solutions doing so...)

Fixed. Thanks for reporting

300 People managed to read the instructions, and it's been a year. Probably not going to be changed. Closing this issue.

I fixed the example test case.

It's solved by +1500 people. Also I just double-checked the kata and it's solvable.

Did you mutate the input array? You shouldn't do that because that affects the test.

Quite possibly, yes, there is something implicit in the instructions.

The requested function needs to have a side effect, instead of returning a value.

This should be apparent from the usage instructions.

Unfortunate typo (?) corrected. Hope that now everything is fine. Lots of thanks for your post.

Thanks for letting me know!

Modified, should work now. If it doesn't, could you tell me? Thanks for your post!