This fundamentally misunderstands both the psychology and the statistics of wallstreetbets type gamblers.

First of all, yes, people are undeniably influenced and motivated by long-shot lottery style payoffs regardless of their expected value or probability of success. Economists have long known, to their great chagrin, that individuals are not perfectly rational actors.

Secondly, it is trivial to manufacture a -EV lottery bet that increases your probability of a desired outcome. For example, suppose that to live “the good life” requires lifetime pre-tax earnings of 5mm. But suppose the median American expects to make half that. By gambling his entire savings on black at the roulette table, with a 47% of winning, he’s making a -EV bet that still dramatically increases his chance of the good life. You can argue that people are incorrectly defining “the good life” and I would agree with you, but that’s a separate argument.

As Russell said, these guys gamble away their futures because they doesn’t believe in their futures otherwise.