There was no bending of the cards, as far as I can tell from this article and another print article I read. The imperfection described on the backs of the cards was related to the way the cards were printed and cut. The other article implied it was something along the lines of the card backs being printed slightly off center relative to where they were cut, so that one long edge of the card might have 2 millimeters of white border between the pattern on the back of the card and the edge of the card, and the other long edge might have 3 millimeters.
As play progressed, Ivey’s associate allegedly told the dealer to rotate certain cards – the ones numbered 6, 7, 8 and 9, which are cards more likely to form a winning hand. Rotated as such, the cards would have the opposite long edge visible in the shoe the dealer deals from. Ivey allegedly then could see (after a pass or two through the deck) whether the first card in the shoe had a 2mm white border or a 3mm white border on the leading edge. Knowing whether this first card had been rotated (to indicate a 6, 7, 8 or 9) would potentially let him know, before the hands were dealt, whether there was a better chance for player or dealer to win… enough to reverse the house advantage.
(Disclaimer: This is my understanding and may be wrong. I read the other article while at work and don’t know which paper it was published in. And the 2mm and 3mm measures I mentioned weren’t explicitly mentioned; I was just trying to illustrate what the article seemed to refer to.)
