If this is worth it to do, wouldn’t it be best for a singular large donor to buy, say, $1M+ of tickets, instead of lots of random individuals buying small numbers of tickets?
Even at scale, the transaction costs would probably negate the positive expected value. Given tickets have to be bought in cash, you would need to get a load of cash out, give it to a load of people, and pay them not very much to buy tickets for you (thus risking they will run off with the cash). I can’t really see how it would take <1 min of person-time per ticket, given the need to also check them, especially if you are picking the numbers manually rather than taking a random draw (e.g. to maximise the chances of not having to share winnings via avoiding common numbers like dates). At +$0.20 per ticket, that is $12/hr!
It’s instructive to look at the limit of buying all the tickets to guarantee winning the jackpot. This has been attempted a couple of times before in other lotteries, with some success. In the Irish lottery in 1992, a syndicate organised by a previous lottery winner, Stefan Klincewicz, managed to buy 80% of the tickets and made a £310k profit on £1.16M of winnings, before expenses. They had people running all over Ireland buying tickets frantically as the Lottery was closing down machines (because it was aware of their plot). It’s unlikely that they made much money at all after paying all these people for their time, and other logistical expenses. And given they only managed to buy 80% of the tickets, the expected value was likely actually negative!
Even if the jackpot grows further and the EV becomes more positive, I imagine this very fact of higher +EV would lead to more people trying such schemes and actually lower the EV (by increasing the chances of a shared jackpot). Maybe this is already happening given the draw being delayed? I don’t think this is a good idea any way you look at it.
If this is worth it to do, wouldn’t it be best for a singular large donor to buy, say, $1M+ of tickets, instead of lots of random individuals buying small numbers of tickets?
Transaction costs negate positive expected value
Even at scale, the transaction costs would probably negate the positive expected value. Given tickets have to be bought in cash, you would need to get a load of cash out, give it to a load of people, and pay them not very much to buy tickets for you (thus risking they will run off with the cash). I can’t really see how it would take <1 min of person-time per ticket, given the need to also check them, especially if you are picking the numbers manually rather than taking a random draw (e.g. to maximise the chances of not having to share winnings via avoiding common numbers like dates). At +$0.20 per ticket, that is $12/hr!
It’s instructive to look at the limit of buying all the tickets to guarantee winning the jackpot. This has been attempted a couple of times before in other lotteries, with some success. In the Irish lottery in 1992, a syndicate organised by a previous lottery winner, Stefan Klincewicz, managed to buy 80% of the tickets and made a £310k profit on £1.16M of winnings, before expenses. They had people running all over Ireland buying tickets frantically as the Lottery was closing down machines (because it was aware of their plot). It’s unlikely that they made much money at all after paying all these people for their time, and other logistical expenses. And given they only managed to buy 80% of the tickets, the expected value was likely actually negative!
Even if the jackpot grows further and the EV becomes more positive, I imagine this very fact of higher +EV would lead to more people trying such schemes and actually lower the EV (by increasing the chances of a shared jackpot). Maybe this is already happening given the draw being delayed? I don’t think this is a good idea any way you look at it.