Hi,

I did not see any mention of that in the description. The 1110000000111 test case is not ambiguous:
since you cannot divide 7 by 3, the only possible speed is 1, while a minimal length algorithm will find that it is 3.
I totally get that you want to orient people toward the minimal length solution for the next kata, and that is what most people who did not think about that case will do. But if you think about that case before coding - as me and probably a lot of people did - then it makes things more complicated.

If you do not want to add this as a test case, then you should add it in the description, stating that the word separator length (0000000) is unreliable to find the speed.

I'm sorry to hear that. On the contrary, it is a very good test case. Algorithms should be efficient and sufficient. The description introduces 5 different operations, but challengers should cover cases those do not include all of them.

111000111 is in standart test set, at least for Python and Java.

I don't consider 1110000000111 to be a good test as it's too ambiguous.

Treating the minimal token length as speed IS a correct way of thinking for this kata and that is explicitly stated in the description. The reason for this is preparing the code warriors for the next, even more complex kata, where this approach is a necessity.