N = 5
board_1 = [[0 for _ in range(N)] for __ in range(N)]

is not the same as:

board_2= [[0]*N]*N

One makes a proper list of lists (the first), the second makes a new kind of animal were board_2[1][1] = 99 changes a whole column.

Oddly, testing board_1 == board_2 is True!

I'm teaching Python to non-majors, and this is a tough one to explain.

This isn't a bug and should be closed. It's more of a stack overflow question.

If you'd like to change this fundamental behavior of a very common operation in python you should make a proposal to the python ideas mailing list at http://mail.python.org/mailman/listinfo/python-ideas

In your example board_2 is equivalent to:

    row = [0] * N
    board_2 = row * N

All the rows are the same initial row. As opposed to board_1 where each row is a new row.

Try this:

    [id(i) for i in board_2]

The initial equivalence is because they do represent the same values (NxN list of all zeroes). What should python compare if not by values?

It's actually fairly easy to explain. Just think about it harder (and consider Yuval's explanation).

http://docs.python.org/faq/programming.html#how-do-i-create-a-multidimensional-list
http://python.net/crew/mwh/hacks/objectthink.html