Thanks for clarification!

Why would team rankings for six teams given in the example come out as [4, 4, 6, 3, 1, 2] if --as the input data shows-- all teams have played a game? Teams 0 and 5 deserve to be in there!
And there was no instruction how to handle instances where a team did not play.
Perhaps as a guy always chosen last in gym class I am missing something here.

Definite mismatch between the terse instructions and the difficulty level unless you are already comfotably using functional languages. Otherwise, once the trick of returning a function becomes clear, it's "So?"

And here I expected a time penalty for brute strength solutions. . .

Frankly, finding something hidden in the test data not truly described in the program requirements really does not prove much apart from tenacity on the part of the programmer, IMHO. It is not that records do not exist for the test case that is tripping up most folks from what I am reading here.