I’m not sure this works, but maybe there’s a way it can. We would need to change the drawing schedule from any week to a much longer interval and probably have to have some cap on the prize size. I’ve looked at some numbers based on the current format and parameters and I get the following:

To create a new address: gas cost to move POOL + gas cost to deposit POOL + gas cost to withdraw POOL is the cost to increase your odds of winning this jackpot. Lets say the user uses 35 gwei with ETH at $2100.

Cost to transfer: $9

Cost to deposit: $18.5

Cost to withdraw: $26

So in total it would cost $53.5 for someone to improve their odds. Each time you try to cheat the system it gets progressively more expensive. It costs $53.5 to double your odds, but the next $53.5 will only improve them by a third… and so on. So each progressive round of cheating becomes less effective.

At the moment the odds for one address to win are 0.087% chance. If cheater spent $642 they would have a 1% chance roughly of winning the draw, assuming others don’t do the same thing, which they probably will. Over the course of the year the would then roughly have a 50% chance of winning the jackpot at least once. If the jackpot is $10,000 that means they spent $642 for a 50:50 chance at winning. For a 5% chance of winning they would need to spend $3,045 and make 58 transactions cycles. So they can spend $3,045 and expect to win the prize at least 2.6 times throughout the year.

Things to tweak:

1). The minimum POOL should be greater than the gas cost to withdraw at a minimum to ensure that the withdraw cost also is a cost of cheating.

2). Extending the gap between drawing schedules will reduce their yearly expectant win rate (i doubt anyone would plan for this being a thing beyond a year, at least right now, might have to address that later).

3). Maybe cap it to 10k or less to disuade cheaters from generating too many tickets.

I haven’t thought about it in great detail, but roughly I think this should be beneficial for fine tuning some idea that isn’t gameable, at least without major inconvenience for the cheaters. One thing that is fair about this is, pretty much everyone that can already use pooltogether is able to cheat.

Edit: I forgot to add the cost to transfer ETH, so the cost is slightly more expensive than I have stated above.