Maybe it's because of my neo-classical economics training, but I'm not like many of my coworkers, who did not have any interest in buying the Mega Millions lottery ticket when the pot was $300 million, all of a sudden couldn't wait to buy their ticket when the pot is $500 million or $640 million or whatever. Like $300 million (even after taxes) is somehow chump change or as if expected value of the lottery isn't the same or much worse than before! Rational economic models can't explain this kind of thing very well. They can't largely because the models are not very realistic. They assume static preferences and preferences based on a perfect calculation of odds into one's utility function- as if each one of us have set utility function that makes any sense!
But in the 1980s as behavioral economics was really starting to gain ground (just after Kahnman and Tversky published their idea called Prospect Theory) To many, expected utility maximization just wasn't sufficient to explain real behavior in decision-making. One of the outputs from that was regret theory. The basic tenet of regret theory is that people aren't just robotic maximizers of their own selfish satisfaction, but more-often in many cases they are regret minimizers.
Regret theory can explain this odd lottery behavior we see in many people today. As the dollar signs skyrocket in unclaimed cash, for a given prices of a ticket the chance of a big regret (of not winning such a historic pot of money) grows significantly, so many people jump into the game to minimize the chances of their regret.