Whale watching on ANV-2 šŸ³

Yeah that feels like a stronger strategy. Any thoughts on how to go about that?

2 Likes

What about a token penalty for those who do not vote? Like 1% of all ANT that have not voted or not delegated their vote gets redistributed proportionally to those who did? Weā€™d vote on that percentage number.

I know delegation isnā€™t there yet, but thatā€™s to say I donā€™t think we should implement that until it is. Like if we force people to vote while they wouldnā€™t have otherwise because they have no idea what to vote, weā€™ll just end up with a mess.

3 Likes

I think that would make more sense for a smaller and more engaged community, but for ANT holders it would cause a riot lol. I do like the idea of creating DAOs that intentionally remove members or reduce their influence if they arenā€™t engaged. It would be really great if there was an app for that with sane defaults so that every DAO didnā€™t have to have a huge conversation around membership and who should or should not be on or off the bus.

I donā€™t necessarily think large ANT token holders would be against, potentially the opposite actually.

It took me a while to get comfortable with the REP use it or lose it for instance, itā€™s quite harsh, but itā€™s also quite effective. As a REP token holder who plans to do my duties, Iā€™m glad this exists as there are multiple positive effects. For one itā€™ll quickly move the value from any token lost back to those who arenā€™t.

I think these types of practices are a bit of a taboo today and Augur broke ground with this, but I expect other projects will follow just because it has great effects.

1 Like

Iā€™m intriguedā€¦ Are there any links/resources youā€™d recommend to learn about how REP goes about implementing this feature specifically and how the community reacted?

Unfortunately I donā€™t have much readings to share about this, sorry. Itā€™s something quite new so at this stage we (the blockchain space) need to experiment and then probably at that point more written feedback will be available.

A better way of taxing would probably be inflation rewards to voters. This seems like a simple and effective solution to get more people voting, and Iā€™m sure it would work in that regard. After all holders are either speculators or users, and the best way to motivate a speculator is through a financial incentive

However @light pointed out that more votes doesnā€™t necessarily mean more voter participation. The simple act of voting doesnā€™t mean you did research, formulated an opinion, and voted.
We could get more votes, but they might be random votes

5 Likes

Yeah thatā€™s a good point. Any mechanism that purely incentivizes voting for the sake of voting does not necessarily incentivize participation in the voting process leading up to the vote. I suggested the Aragon Voting Gauntlet as a way to make the process more involved from the perspective of people submitting proposals so that they would have to engage with the community and couldnā€™t just submit things last min. Itā€™s not perfect, but if more voters are engaged leading up to the vote it could result in better outcomes without needing to monetarily incentivize anyone?

2 Likes

The mechanism I would suggest is exactly like the one @Fran23 pointed out. One way to
mitigate the random voting problem would be to also distribute rewards to voters who predicted correctly on the outcome (quite nominal like 0.5% of total votes submitted), but not incorporate any sort of slashing mechanism for being wrong.

That way participation always nets you more influence over time. However, without proper tools to identify voters and their behaviors, there is an easy attack vector in which small cliques of large ANT holders form to consistently influence decisions, making it even more gameable.

4 Likes

Letā€™s test it. Without even changing token logic or creating an inflation mechanism. Just using AA funds for the test drive.

Next voting cycle, we create a new AGP going along these lines:

Q) Should people voting this AGP get a reward from the Aragon Foundation, equal to x% of the tokens used to vote?
A)
Yes -> people get tokens
No -> noone gets nothing but we attracted lots of attention to the process

Would be interesting to see how easily people are motivated to participate, without getting random votes on the other AGPs

I would probably vote no fwiw :grinning:

I would vote against rewarding voters for voting, from inflation or from Aragon funds. Only thing I would personally accept is to lightly slash (punish) non-voters. And again, only when delegation is possible so people who donā€™t know what to vote for, can rely on more knowledgeable people they trust.

edit: only place Iā€™m willing (eager) to reward voters for voting is once a proper way to do it is implemented: Futarchy.

@gh1dra
Actually you could impliment that as a prediction market separate from Aragon right? Since all the votes are on-chain, you could go to the predication market with the same account you used to vote and then collect a reward.
That being saidā€¦ who provides the reward? AA could provide a reward that only goes to the winnder, but then that motivates people to vote for what they think will win, not necessarily what they actually want to win. This gives whales more influence.
Thereā€™s 100% nothing stopping a whale (or anyone) from setting up a rewards app where if your account voted for X on vote# Y, you get a reward. Today itā€™s socially taboo to do so, but itā€™s probably only a matter of time until we start seeing these types of games everywhere.

@Fran23

Would be interesting to see how easily people are motivated to participate, without getting random votes on the other AGPs
That would be interesting. Thereā€™s nothing to lose by just submitting the AGP and watching the debates ensue. Iā€™d actually be really curious to see that happen, not because of the outcome, but because Iā€™m sure a lot of smart people would weigh in and Iā€™d learn a lot lol

@GustavMarwin
Good point. Delegation is going to be an essential component to making this whole process productive rather than chaotic

2 Likes

Definitely agree that voter rewards doesnā€™t necessarily lead to a more informed vote.

As @GustavMarwin just suggested, Futarchy is what we seem to be converging on. Although there is the question of who gets to decide the metrics of success prior to running the elections for future AGPs. If we were to use the current voting apparatus, it may produce an even more ā€œplutocraticā€ outcome. Excited to see this app launch so we can run experiments :test_tube: :lab_coat:

Iā€™d love that, but there are several issues:

  • Slashing tokens may hurt users that have them in exchanges, and basically make life impossible for exchanges. Imagine explaining to a user that their balance has been reduced in the meantime between depositing and putting an order
  • Same with tokens locked on a contract

Inflation has the same properties of slashing, since it increments value for those who receive more tokens, and decreases it for those who donā€™t.

The problem is, of course, random voting.

However, I have this WIP idea around token locking.
You could apply voting multipliers to voters who lock their tokens for a period of time. For example, you lock your tokens for 4 years and get 2x voting power
As time goes on and the lock is reduced, you decrease the multiplier. For example, 2 years in you now have a 1.5x multiplier. You can re-lock to get it up again

This will end up making it so the votes are mostly dominated by people who are long Aragon, and have skin in the game in the long term of the project.

With that, you can reward them with some small inflation, since a rational voter that is locked for X years wouldnā€™t vote proposals that hurt the project, and therefore risk the value of their ANT.

Iā€™d love to arrive to a day in which there are entire teams of people analyzing each proposal and voting for the best interest of the Network, and even making a living out of it. That would crystallize with futarchy in the long term.

12 Likes

Interesting idea. Once my tokens are locked for 4 years, would I be able to use them to pay for the network fees once they will be implemented? My current understanding is that those would only be voting tokens during that period, which has its disadvantages. Would it be possible to simply look at how long an address has kept its balance or would that not work?

As for the inflation vs slashing, itā€™s an interesting point. I think anyone that started early on the blockchain space has a bias for no-inflation due to the Bitcoin culture. Overall what we might need to focus on individually is how our share of the network is evolving; if thereā€™s inflation but our share remains the same, itā€™s really what matters. I guess thatā€™s the thinking the Aragon project has been pushing from the start.

1 Like

I think that should be the case, or even put them as collateral for Aragon Network agreements.

The issue with that is that they are still liquid, so you can vote on a decision and then dump all of them. With locking, we want to prevent that exact scenario

Exactly! And with this system, if you participate on it, your share may even grow!

2 Likes

Some thoughts on time-lock voting and other topics that came up in this threadā€¦ this post turned into a bit of a jumble of different things. :upside_down_face:

wrt time-lock voting mechanism add an additional layer of ā€œskin in the gameā€ to stake-weighted voting.

There are a few ways to do this:

  1. Only allow tokens which are locked for a certain amount of time to participate in the voting process.
  2. Allow all tokens to vote, but provide more weight to tokens which are locked. (this could be done based on a step or continously, but should always converge to some max multiplier regardless of length of time submitted, otherwise burning a single token gives you infinite voting weight).
  3. Make voting weight = locking duration. This means that in order to support a proposal you must lock tokens t for some fixed duration, d, and your vote weight is t * d for any proposal. If you apply a non-linearity to this, could have similar advantages to Quadratic Voting (though any non-linearity here would introduce advantages to splitting tokens between accounts and likely require some sort of sybil resistance if there decisions have significant consequences.

Itā€™s worth noting that all of these make the cost of participation higher. This isnā€™t a bad thing, we are not trying to optimize for participation rates but rather good decisions but since this thread is in response to the idea that a single whale was able to cast a decisive vote, a system like this may exacerbate the issue of relatively small amount of stake being able to be decisive.

I think it is also worth noting that due to the current state of ANT liquidity there is already a pretty significant ā€œlockupā€ effect inherent in holding ANT that is not directly related to voting. This effect makes it unlikely to see large amount made available on defi lending platform, and makes large holders especially, have a strong motivation to pursue policies that create long term value.

The main reason Iā€™m skeptical of introducing a time-based locking mechanism is that it further reduces the ability for participants to exitā€“and without things like the court which help enforce rights of minority participants, the ability to easily exit is a really important check on the governance process. I would be much more comfortable with a time-based lock in requirement if once we have the court to serve help with dispute resolution.

Unrelated to the idea of time lock voting, I think that introducing a ā€œquiet endingā€ policy or simply switching to commit/reveal process for votes would help mitigate the (potential) issue with waiting till the end of a voting period to swing votes (or voters in general choosing to vote or not based on the current state of the vote before it ends).

Also with regard to incentivizing voting participation, I think that we 1) definitely need to be careful about how this is designed because as was noted above it can result in random votingā€¦ combining it with a time-lock up mechanism can help address this concern but its still not clear to me if participation rate is something we really want to try and optimize forā€¦ 2) If implemented via inflation there may be some concerns with ā€œtax liabilityā€. Iā€™ve been a big fan of inflationary participation rate targeting mechanism from a theoretical perspective, but from a practical perspective it may be better to implement as Demurrage even if it is significantly more technically complicated. Some discussion on this issue for more context.

3 Likes

I agree and this is essential to keep in mind.

Where can one find more info on the Court, specifically how it will help enforce the rights of minority participants?

Thatā€™s a really good point and one that isnā€™t discussed enough. The mechanism described on the Cosmos forum wouldnā€™t really work for Aragon though because thereā€™s no staking. Is there a different type of mechanism you were thinking of for Aragon?

1 Like

Checkout the recent update to GitHub - aragon/whitepaper: An opt-in digital jurisdiction for DAOs and sovereign individuals

also Requirements for Aragon Network / Court Launch - #2 by burrrata

And have a backlog of additional posts/documentation in my task queue so expect more info to come out in the next few weeks. :slight_smile:

I think if we enforced some sort of time-lock voting and some sort of penalty for locking and not voting, it would actually closely resemble staking.

2 Likes

That could have an interesting affect on market liquidity. Theoretically the value of ANT would be derived from itā€™s usefulness (ANV, Court, etcā€¦), but it would only be useful if it was staked/locked to do that stuff.

  • In the short term this would create an opportunity cost of selling/exiting because it selling would reduce voting/participation power. This might further constrain supply and drive up the price as more people want to do stuff with ANT, but thereā€™s less and less ANT to buy. On the other hand, markets reflect future expected prices so traders might just keep their tokens liquid and price them reflecting expected future utility regardless of the current discount that results from the token not being staked/locked.
  • In the long term though atomic swap mechanisms will allow people to stake/lock ANT, but still exchange the keys that control them without the blockchain knowing the difference. This would kind of ruin the whole staking/locking game.
1 Like