The anti-bot deposits increase the cost of attacking the POI network. The bots need to control a majority of a group to be able to verify one another, the pseudonym events use a point-system where 3 nyms always win over 2 nyms and over 1 nym, so for every group bots control, there would be lots of 1 and 2 account groups that they would loose their deposit in.
The cost to attack the POI network can be measured in depositPenaltiesPerPOI
, the deposits for all bot accounts that end up in groups that are not under the bots control. This cost would have to exceed to value of a POI, if they are to prevent attacks.
If POIs are used for crypto basic income, for example swarm redistribution, and a basic income is around $1111, then depositPenaltiesPerPOI * depositSize
would need to be higher than a basic income of $1111.
The ratio of groups with a majority of bots seems to decrease when the total number of users increases, if the percentage of bots is fixed.
Running a couple of tests,
int groupSize = 5;
int majority = 3;
with different population sizes, int population
, and
int bots = population / 10;
int numGroups = population / groupSize;
int groups[numGroups];
int groupControlled[numGroups];
int groupsControlledByBots = 0;
and assigning the bots into groups by random,
int rand_max = 2147483647;
int denominator = rand_max / numGroups;
int a;
for( a = 0; a < bots; a = a + 1 ){
int groupNumber = rand() / denominator + 1;
groups[groupNumber] += 1;
if(groupControlled[groupNumber] == 0) {
if( groups[groupNumber] > majority ) {
groupsControlledByBots += 1;
groupControlled[groupNumber] = 1;
}
}
}
gave the distribution in the graph above, here it is again,
So if I understand this your saying an attack on a PoI setup would be possible but not profitable?
Downvoting a post can decrease pending rewards and make it less visible. Common reasons:
Submit