It is not mandatory to test all the possible outputs in fixed tests.

No test cases for [queen, bishop] solutions: example of one:

`
test case
test.assert_equals(

promotion(
  [
    [' ', ' ', ' ', ' ', ' ', ' ', ' ', ' '], 
    [' ', ' ', ' ', ' ', ' ', ' ', ' ', ' '], 
    [' ', ' ', ' ', ' ', ' ', ' ', ' ', ' '], 
    [' ', ' ', ' ', ' ', ' ', ' ', 'K', ' '], 
    [' ', ' ', ' ', ' ', ' ', ' ', ' ', ' '], 
    [' ', ' ', ' ', ' ', ' ', ' ', ' ', ' '], 
    [' ', ' ', ' ', ' ', ' ', ' ', ' ', ' '], 
    [' ', ' ', 'P', ' ', ' ', ' ', ' ', ' ']
  ]
), 
["queen","bishop"]

)

`

The rule to break tie should follow classic piece ranking

Queen, rook, bishop, knight in this order

Am i wrong?

I don't know if you are wrong. It may be a good exercise but there are more straightforward approaches. The input is just a 8 * 8 array.

You are on the wrong track. No need no graph theory to solve it.

I think this is terrific graph problem but I'm not sure it belongs on the current difficulty it is at :) (currently 6). Anyone else feel the same?