We now measure position and speed relative to the portals, which improves readability. But how do you explain the remaining jump in speed? We saw earlier that a jump in speed happens when the cube hits something hard, but here, there is nothing in its way... Maybe our assumption was wrong?