Yes, I think you guessed right. If f is an empty filter object (list(f) == []), then in Python 3 bool(f) returns True which is why you do not get what you expected.

Third line checks if there are any columns, rows or blocks which are not a set of numbers from 1 to 9.
Second line helps define blocks.
Does this help you? You can ask more specific questions if you are missing something.

the %s are placeholders
the string is formatted by replacing the placeholders with the corresponding values from the tupel.

Actually, that was the sample tests. Here is something for the first fixed test:

game word is: aver

Alice : aver to over
Bob   : over to oven
Alice : oven to oxen
Bob   : oxen to ....

and okay, why not, here is the entire first set of fixed tests:

alice = ["fare", "cans", "left", "slap", "gain", "slam", "pate", "even", "mane", "peel", "flat", "oxen", "claw", "meet", "mats", "pulp", "over", "faze", "goat", "gone"]
bob   = ["oats", "gaze", "feet", "feat", "pats", "golf", "cope", "maze", "caps", "flaw", "boot", "slaw", "pare", "oven", "mine", "clap", "goof", "cane", "fool", "oast"]
test.assert_equals(mutations(alice[:], bob[:], "aver", 0),  0) # Alice goes  first, Alice   wins
test.assert_equals(mutations(alice[:], bob[:], "spam", 0),  1) # Alice goes  first, Bob     wins
test.assert_equals(mutations(alice[:], bob[:], "gulf", 0),  1) # Alice fails first, Bob     wins
test.assert_equals(mutations(alice[:], bob[:], "wren", 0), -1) # Alice fails first, neither wins
test.assert_equals(mutations(alice[:], bob[:], "maps", 1),  1) # Bob   goes  first, Bob     wins
test.assert_equals(mutations(alice[:], bob[:], "oars", 1),  0) # Bob   goes  first, Alice   wins
test.assert_equals(mutations(alice[:], bob[:], "moat", 1),  0) # Bob   fails first, Alice   wins
test.assert_equals(mutations(alice[:], bob[:], "kiln", 1), -1) # Bob   fails first, neither wins

Hi, the breakdown for this test is found in the kata description:

3. In the case of word = "maze" and first = 0:
* Alice responds to maze with mare
* Bob responds to mare with more
* Alice responds to more with pore
* Bob responds to pore with pure
* Alice responds to pure with pare
* Bob has no valid response to pare
* Alice wins the game.

Hope that helps, best of luck!