This strategy makes sense when the odds of winning are known, and the jackpot is not shared.
The powerball jackpot is shared so you would need to factor in the likely number of people you would share with which would be complicated to figure out.
There are cases in gambling where the jackpot is not shared, for example poker machines.
The odds of winning remain fixed, yet the jackpot slowly increases, moreso when there are many machines linked to build the same prize.
At a certain level of jackpot the math works out and it's a sound strategy to pour money into the machines.
If you were going to invest in lotto with shared jackpots I've heard anecdotes that choosing numbers 32 and higher is a better way to go because people tend to pick birthdays. I think looking at past drawing results and whether there was a winner or not you could analyse if numbers are picked disproprotionately, but I'm not sure what sample size would be needed to generate meaningful results.
All that said, I can think of many things I'd rather develop expertise in than reverse engineering the payout odds of poker / progressive jackpot machines.
The powerball jackpot is shared so you would need to factor in the likely number of people you would share with which would be complicated to figure out.
There are cases in gambling where the jackpot is not shared, for example poker machines.
The odds of winning remain fixed, yet the jackpot slowly increases, moreso when there are many machines linked to build the same prize.
At a certain level of jackpot the math works out and it's a sound strategy to pour money into the machines.
See: break even point https://en.m.wikipedia.org/wiki/Progressive_jackpot
Here's a person who's written about dedicating their life to this approach: http://www.amazon.com/Million-Dollar-Slots-Peter-Liston/dp/0...
If you were going to invest in lotto with shared jackpots I've heard anecdotes that choosing numbers 32 and higher is a better way to go because people tend to pick birthdays. I think looking at past drawing results and whether there was a winner or not you could analyse if numbers are picked disproprotionately, but I'm not sure what sample size would be needed to generate meaningful results.
All that said, I can think of many things I'd rather develop expertise in than reverse engineering the payout odds of poker / progressive jackpot machines.