Yes, my approach lessens the voting weight of accounts that share similar voting patters. I posted a walkthrough of the methodology a while ago but I guess it flew over people's heads.
votingweight = ( 1 similarity ) ^ (4/3) * (accountbalance)^(1/3) * (accountage)^2 (timesvoted)^(3/2)
or something along these lines. "similarity" increases as accounts make the same voting choices as other accounts.
timesvoted= amount of times that account number has voted
timesvoted prevents people from creating additional accounts to circumvent the similarity variable
Similarity is very easily defeated by voting randomly on issues which doesn't concern the real voter.
You're welcome to try and solve this enigma with me
.
Realistically, it would be such a phenomenal pain in the ass to game the system that I doubt anyone would try.
Real world voting system is also far from perfect. So let's be constructive and test the system and improve it gradually.
We did many polls before such as rewards, logo etc., which can be subject to multiple type of manipulation. But then no one had any complaint about the result and we can move on to other things.
Here is one of the poll I did earlier
https://bt.irlbtc.com/view/423241.0 and the result looks pretty legit to me. You are welcome to vote there or manipulate it if you like. It should be clear that the poll is not authoritative but we can still use it as an anchor to move forward if no one has a strong objections.
If we want to get things done, we should do things pragmatically. I would like to see coolmist's system being tested in a complete form to see how it differs to simple poll results.
