[VoterApathy] How To Make 51% Attacks A Lot Harder (without more people voting)

[xeroc]

All you need is the ability to make "voting power" expensive.
In Bitcoin, you need to buy expensive equipment
In BitShares you either need to buy stake or convince people to put trust into you and set you as a proxy.

What would buying 16% stake result in? Is there even 16% available in exchanges (for sale)?
Wouldn't that also lead to passive shareholders to claim their stake and read into voting. Maybe maybe not.
We will see sooner or later what will happen when the price goes up (for whatever reasons)

Alternatively, how would you convince people to trust you with their voting power? Isn't paying for votes perceived
as corruption? Would you vote for a president that pays you $20 right away? Do you think that people with a lot of
stake (face in the game) would take the offer as well?
Why would I want to give control over to someone that I do not trust, that may do no good to the value of my shares?

Any finally, if shareholders are unsatisfied by the voting distribution, it is in their hands to change it. Similar to what
Bitcoin has seen with GHash.IO.

Furthermore, Just because the proxies have a lot of voting power does NOT necessarily mean that will use it to replace
witnesses and even if that happend, all they can do is halt the network.

In the end, BitShares is way more flexible when it comes to governance even though it may be easier currently to get
dictatorship. But IMHO, it is a good thing that @bytemaster can stear the boat (for the moment) and protect it from
hostile take overs. What we have seen in BitShares 1 already is that someday, he will NOT have the power anymore to
stear the boat alone.

[Empirical1.2]

The 0.13% BTS attack

Maybe you don't need 13% stake to attack BitShares maybe you only need 0.13%...

As a proxy offer to pay 1% interest per month to accounts that proxy their stake to you. That would be a lucrative 12%+ p.a

1 months interest payments on 13% of proxied stake would only cost you 0.13%...

Except that if everyone did this BTS would fall more than 1% and thus no one would do it. Hence, it is not profitable for people to proxy.  Now if you offered them 10% and the attack did less than 10% harm to BTS price then it would only cost 1.3% but even that is likely not profitable for the attacker because the network would quickly recover and the attacker would be out $100K or more.

I just posted this as a separate topic but I can't delete my own threads apparently.

Yes but your response supposes shareholders would know it was an incentive being offered for the purpose of an attack whereas it could be masked via multiple proxies and different intentions.

[bytemaster]

The 0.13% BTS attack

Maybe you don't need 13% stake to attack BitShares maybe you only need 0.13%...

As a proxy offer to pay 1% interest per month to accounts that proxy their stake to you. That would be a lucrative 12%+ p.a

1 months interest payments on 13% of proxied stake would only cost you 0.13%...

Except that if everyone did this BTS would fall more than 1% and thus no one would do it. Hence, it is not profitable for people to proxy.  Now if you offered them 10% and the attack did less than 10% harm to BTS price then it would only cost 1.3% but even that is likely not profitable for the attacker because the network would quickly recover and the attacker would be out $100K or more.

[Empirical1.2]

The 0.13% BTS attack

Maybe you don't need 13% stake to attack BitShares maybe you only need 0.13%...

As a proxy offer to pay 1% interest per month to accounts that proxy their stake to you. That would be a lucrative 12%+ p.a

1 months interest payments on 13% of proxied stake would only cost you 0.13%...

[Empirical1.2]

You would incentivize individuals to vote but not for any one particular proxy/initiative thereby getting them into the polling box where the majority should vote for the initiatives they actually agree with.

I think it's very unlikely that people who currently can't be bothered to select a proxy, will suddenly actively seek information and make informed voting decisions.
You'll have to pay a lot to change their attitudes so much. Convincing them to use proxies is much easier and cheaper and the final effect is almost the same, or even better as we avoid random voters.

If a beauty pageant wanted to get local town people to come in and vote for the prettiest girl, would they be better off...

A) Give $1000 to every voter that turns up and votes
B) Give $1000 to each girl for every guy they get to vote for them

Even if voters were lazy, I would say that A would increase voting but the prettiest girl would still win.
I would say that B would increase voting but the girl who passed the most money onto voters and had the best marketing of that voting reward may be more likely to win.

So B could give you a bad result. (Obviously replace prettiest girl for stronges initiative or best proxy.)

A very inadequate analogy. Proxies don't run against each other like those girls, they just facilitate the process of voting. 
It's a service that makes voting easy. Just like a bus makes moving around easier than walking on foot.

We are currently straggling to make people do a very simple thing - select a proxy.
According to you, we should aim to convince people to practice regular jogging while we are currently unable to make them even stand up.
Let's first learn to crawl before we attempt to run - as BM recently said.

I don't think BTS being controlled by a handful of proxies will be great for marketing, even though higher voting numbers would be good. However I do agree with you that individual voting is harder than selecting a proxy.

I still think incentivising voters directly, (even if it's too select a proxy) is better than incentivising the proxies to get votes, because the first would get voters to select the best proxy while the second would get voters to select the best marketed &/or paying proxy.  Even though you found my analogy to explain why inadequate.

[jakub]

You would incentivize individuals to vote but not for any one particular proxy/initiative thereby getting them into the polling box where the majority should vote for the initiatives they actually agree with.

I think it's very unlikely that people who currently can't be bothered to select a proxy, will suddenly actively seek information and make informed voting decisions.
You'll have to pay a lot to change their attitudes so much. Convincing them to use proxies is much easier and cheaper and the final effect is almost the same, or even better as we avoid random voters.

If a beauty pageant wanted to get local town people to come in and vote for the prettiest girl, would they be better off...

A) Give $1000 to every voter that turns up and votes
B) Give $1000 to each girl for every guy they get to vote for them

Even if voters were lazy, I would say that A would increase voting but the prettiest girl would still win.
I would say that B would increase voting but the girl who passed the most money onto voters and had the best marketing of that voting reward may be more likely to win.

So B could give you a bad result. (Obviously replace prettiest girl for stronges initiative or best proxy.)

A very inadequate analogy. Proxies don't run against each other like those girls, they just facilitate the process of voting. 
It's a service that makes voting easy. Just like a bus makes moving around easier than walking on foot.

We are currently straggling to make people do a very simple thing - select a proxy.
According to you, we should aim to convince people to practice regular jogging while we are currently unable to make them even stand up.
Let's first learn to crawl before we attempt to run - as BM recently said.

[jakub]

Before we discuss any other means, why don't we start with the simplest thing available: let's encourage the users to vote in the GUI.

We can do it in three phases:

(1) If the system finds out you are not voting, it should display a red warning in the footer (just as we have it for the wallet backup reminder).
Voters apathy is a serious vulnerability for the blockchain, and I guess most people don't vote because they don't realize that.
And they don't realize that because we did not bother to tell them this in the first place.

(2) On the voting page there should be a clear instruction that the easiest way to vote is to select a proxy and offer a link to this forum board.

(3) We could enable a special flag associated with an account which would indicate that this account has officially declared to be a proxy.
This way we could have a list of proxies in the GUI (as we have the list of witnesses) and users would be able to choose a proxy even without referring to the forum.

EDIT: I've made a github post:
https://github.com/cryptonomex/graphene-ui/issues/630

[Empirical1.2]

Incentivising proxies might not be ideal as even though you may end up with more stake voting, which is a positive,  you will end up with very large proxies which is bad for marketing.

How did you arrive at the conclusion that we will end up with large proxies?
Isn't it more likely that we will end up with a large number of proxies?
It all depends how we structure the financial incentive] - e.g. we can pay more for the first 10M BTS of followers and less for the next 10M.

In that example the Proxy could make multiple accounts.

I would suggest the financial incentive should be even smaller, 100 000 BTS to encourage individual voting.

Even now we end up with a very small number of proxies, which centralizes power and makes BTS appear centralized. You could argue that if one of the proxies turned bad people would remove their support but this is the same as hoping hashers will leave a bad pool. Better to disincentize large pools in the first place imo & it's much better for marketing.

You would incentivize individuals to vote but not for any one particular proxy/initiative thereby getting them into the polling box where the majority should vote for the initiatives they actually agree with.

I think it's very unlikely that people who currently can't be bothered to select a proxy, will suddenly actively seek information and make informed voting decisions.
You'll have to pay a lot to change their attitudes so much. Convincing them to use proxies is much easier and cheaper and the final effect is almost the same, or even better as we avoid random voters.

If a beauty pageant wanted to get local town people to come in and vote for the prettiest girl, would they be better off...

A) Give $1000 to every voter that turns up and votes
B) Give $1000 to each girl for every guy they get to vote for them

Even if voters were lazy, I would say that A would increase voting but the prettiest girl would still win.
I would say that B would increase voting but the girl who passed the most money onto voters and had the best marketing of that voting reward may be more likely to win.

So B could give you a bad result. (Obviously replace prettiest girl for stronges initiative or best proxy.)

[jakub]

Incentivising proxies might not be ideal as even though you may end up with more stake voting, which is a positive,  you will end up with very large proxies which is bad for marketing.

How did you arrive at the conclusion that we will end up with large proxies?
Isn't it more likely that we will end up with a large number of proxies?
It all depends how we structure the financial incentive - e.g. we can pay more for the first 10M BTS of followers and less for the next 10M.

You would incentivize individuals to vote but not for any one particular proxy/initiative thereby getting them into the polling box where the majority should vote for the initiatives they actually agree with.

I think it's very unlikely that people who currently can't be bothered to select a proxy, will suddenly actively seek information and make informed voting decisions.
You'll have to pay a lot to change their attitudes so much. Convincing them to use proxies is much easier and cheaper and the final effect is almost the same, or even better as we avoid random voters.

[Empirical1.2]

How about rewarding  proxies based on stake voting through them?
Currently it's a hard and thankless task keeping up with all the witnesses and staying on top of all the proposals.
Why not make it someones BUSINESS to promote voting?

 
We are a financial platform so financial incentives are the most straight-forward tool available.
The best option is to incentivize proxies to actively look for followers - this way we will achieve our goal without the downside of getting random votes.

There seems to be no consistency in our behavior: we say we want something (which indicates we find it valuable) but at the same time we say we don't want to pay for it (which indicates we don't find it valuable).

Incentivising proxies might not be ideal as even though you may end up with more stake voting, which is a positive,  you will end up with very large proxies which is bad for marketing. The power distribution (perceived centralisation) will look worse than Bitcoin https://blockchain.info/pools?show_adv=no

Also whether you vote for proxy X will then be determined by how much of their earnings they pass on and how good they are at marketing themselves, vs. whether you think proxy X best serves BTS from your perspective.

Whereas if you  awarded the the proceeds of the 5% annual inactivity fee on a monthly basis with rewards limited up to the first 100 000 BTS in that account...

You would incentivize individuals to vote but not for any one particular proxy/initiative thereby getting them into the polling box where the majority should vote for the initiatives they actually agree with.

Even if people split up their stake into smaller amounts to gain maximum benefit, BTS will get the marketing benefit of more stake voting frequently and the power distribution graph (Size of stake in each account - perceived centralization) will be far more distributed as well.

[jakub]

How about rewarding  proxies based on stake voting through them?
Currently it's a hard and thankless task keeping up with all the witnesses and staying on top of all the proposals.
Why not make it someones BUSINESS to promote voting?

 
We are a financial platform so financial incentives are the most straight-forward tool available.
The best option is to incentivize proxies to actively look for followers - this way we will achieve our goal without the downside of getting random votes.

There seems to be no consistency in our behavior: we say we want something (which indicates we find it valuable) but at the same time we say we don't want to pay for it (which indicates we don't find it valuable).

[Empirical1.2]

There are people who are passionate enough about BitShares to vote and do work to establish a reputation that others trust to vote for them.

What do we gain if we water down that trusted group with others who are merely paid to think for us?  How do we know that a paid voter is wise and actually doing her homework?

The average knowledge/passion/trustworthiness per vote would decrease.  Why would that be good?

Anything done to artificially increase the number of votes cast without increasing the knowledge behind those votes is just for show.  Which means that the %voting metric is a poor indicator.

Suppose we offered $100 per share to get people to vote just one time.  Naturally, turnout reaches 99%.  What have we done?  We have now locked in a bunch of random noise that averages out to the same thing plus a bias based on the sum of all disinformation, rumors and misconceptions.  Now, when conditions change, the resulting inertia of unattended one-time paid votes is much harder for the truly informed and motivated voters to make a change.

Isn't part of the premise of FMV that paying people to vote is beneficial at times?

Your example also references a one off voting payment not a continuous voting incentive.

A person with >5 million BTS  might feel he can make an impact and increase the value of his considerable stake by voting. A person with <100 000 BTS may feel neither of those but in order to make BTS more marketable than being susceptible to a 13% attack we would benefit from those sub 100 000 BTS holders voting. (Even though they may own stakes in 15 other projects and it's not currently worth their time/effort to vote.)

Many people don't bother to vote in local elections but if they were paid to turn up, they would. The majority wouldn't vote randomly but in favour of the parties and ideas they already support imo. So an incentive for frequent voting may actually make BTS more secure and not just be noise.

[triox]

There are people who are passionate enough about BitShares to vote and do work to establish a reputation that others trust to vote for them.

What do we gain if we water down that trusted group with others who are merely paid to think for us?  How do we know that a paid voter is wise and actually doing her homework?

The average knowledge/passion/trustworthiness per vote would decrease.  Why would that be good?

Anything done to artificially increase the number of votes cast without increasing the knowledge behind those votes is just for show.  Which means that the %voting metric is a poor indicator.

Suppose we offered $100 per share to get people to vote just one time.  Naturally, turnout reaches 99%.  What have we done?  We have now locked in a bunch of random noise that averages out to the same thing plus a bias based on the sum of all disinformation, rumors and misconceptions.  Now, when conditions change, the resulting inertia of unattended one-time paid votes is much harder for the truly informed and motivated voters to make a change.

I'm talking about paying proxies. Let them figure out how to  mobilize voters.

[Stan]

There are people who are passionate enough about BitShares to vote and do work to establish a reputation that others trust to vote for them.

What do we gain if we water down that trusted group with others who are merely paid to think for us?  How do we know that a paid voter is wise and actually doing her homework?

The average knowledge/passion/trustworthiness per vote would decrease.  Why would that be good?

Anything done to artificially increase the number of votes cast without increasing the knowledge behind those votes is just for show.  Which means that the %voting metric is a poor indicator.

Suppose we offered $100 per share to get people to vote just one time.  Naturally, turnout reaches 99%.  What have we done?  We have now locked in a bunch of random noise that averages out to the same thing plus a bias based on the sum of all disinformation, rumors and misconceptions.  Now, when conditions change, the resulting inertia of unattended one-time paid votes is much harder for the truly informed and motivated voters to make a change.

[triox]

A well informed vote from someone who has done the hard work to understand the issues is highly valuable.

A random (e.g. machine generated) vote is useless - it just increases the noise floor without changing the outcome.

Alas, rewarding voting yields mostly the latter, not the former.

How about rewarding  proxies based on stake voting through them?
Currently it's a hard and thankless task keeping up with all the witnesses and staying on top of all the proposals.
Why not make it someones BUSINESS to promote voting?