You’re right that choosing less-frequently-chosen numbers (and, in particular, avoiding the numbers that can appear in dates) improves the expected payout by making it less likely that you’ll share the prize if you do win. But the ABC story didn’t recommend that. They recommended betting on automatically-chosen numbers.
Their recommendation is faulty not only because a popular number is as likely to be auto-picked as a unpopular one, but also because their “reasoning” went like this: most winners auto-pick, therefore there is some unknown mechanism that makes autopicked entries more successful. In fact, there is no such mechanism, and auto-picks win more simply because they’re played more.
Watch the video. It’s a cluster-fuck.
