no specifications about equal ranks

The issue below is not fixed: stringified objects are still expected to have a specific order of properties. Tests should be updated to Node14 and Chai assertions, because chai.assert.deepEqual handles object equality correctly.

returning string representation now

Randoms test cases added.

The tests reject return values that would deep-equal the expected values. Solutions are currently expected to return values with a string representation that equals the expected ones'. These tests currently rely on undefined behavior: The ordering of JavaScript objects' properties is not defined.

the first 0 argument is not necessary

Check my new solution - random test cases are needed to prevent basic form of cheating from users, all new katas unless it's a specific type of a puzzle kata should have random test cases, I won't be able to approve it otherwise.

It already has test cases. Randoms test cases are not mandatory.

Needs random test cases.

thanks for your aclaration

Points should be zero after new ratings are calculated.
"gains of loss" should be "gains or losses"

it doesn't mention that rank should be updated according to order.

Kata instructions don't mention any changes to Rank. Would be clearer to mention this up front rather than relying on failed tests.

while I agree, if you're displaying the result on an scoreboard. the points that caused the change in the rank are useful.