Anchor's Poll function can do better

Community Initiatives
1

After experiencing the poll function of Anchor, and comparing it to other protocols, there are some glaring shortcomings of what we have here at Anchor.

  1. The cost to start a poll is too high, therefore there are no suggestions being put forth.

This has been voted on and the poll failed to reach quorum. I feel terrible for whoever sponsored this, because it made a lot of sense to pursue, yet nobody gave a dam, if they didn’t agree with it, they could have voted no. But nobody voted.

The solution would be to revisit the idea of reducing the poll cost, perhaps 100 ANC might be too low, I was personally thinking 300 ANC would have been easier to swallow.

  1. Everyone benefits if a good idea is executed. but the sponsor of the poll has no direct benefit, yet has to risk loosing the ANC if there is a lack of quorum.

The poll sponsor should be rewarded for the undertaking the risk of starting a poll, possibly the doubling of poll stake, and have better measures for rejecting a poll, rather than just not reach quorum, of which 10% does seem rather high to be reached, especially if we have multiple polls that need to be run at the same time, and seeing that nobody is really incentivised to respond to polls these days, because ANC is constantly loosing value, nobody will want to hold onto ANC just to vote. See next problem…

  1. Sometimes Polls need more than one answer.

A binary outcome might not suit everyone who would like to vote and the only way they can participate is therefore to not participate.

Solution: Create the option for multi optioned polls (Maybe up to 3 choices?), create the option for yes, no, alternative choices, abstain and reject.

The abstain option will people to vote when they feel they aren’t informed enough to make a decision and will allow for participation even when you don’t agree with a possible binary outcome.

The reject option will be for people who want to protest truly terrible ideas being put forth, and this will be the measure if a poll is rejected and the poll’s staked ANC refunded to the poll sponsor. It should be applied that people who reject a poll, will receive their governance staking rewards, but will NOT receive the share of the poll stake, instead their share will be burnt or sent to the community pool/treasury. Poll sponsors who keep making useless polls with the hope that people will pass their poll and be rewarded will therefore be subject to this.

  1. Nobody bothers to vote and the rewards for participating in governance are terrible.

This one is easy, currently it’s terrible holding ANC because it’s deflationary, we should therefore compensate people for who actually help the governance process. ANC governance rewards should be therefore double just for being staked towards governance and tripled for the short time they are actually in use when voting. I would imagine having a 50-70% Apr for the couple of weeks there are things to vote on is actually should be the norm in the developing stages of the protocol. 25% APR is nowhere near enough for ANC holders to hold onto ANC in the face of the massive deflation that is occurring as a result of borrower incentives being sold into the LPs to cover their cost. (This is actually another massive reason why the sufficiently large volume LPs are actually strategically VERY important, so as to dampen the sell pressure of the borrower incentive ANC trading back into another currency)

1 Like
2

Can we get another protocol as a comparison for what you say is better?
I know of snapshot votes but those are different beasts, TEXT proposals aside, Anchor Governance is designed around direct execution, voters pass the vote and it gets executed, snapshot is designed around gathering information to later be executed if the multisigs desire, not exactly prime decentralization.

Also that proposal that missed quorum, it was a shame it didn’t reach quorum but I’m glad it did not pass, if you pay any attention to the Mirror Governance (which passed a similar proposal) it was a target of a governance attack, filled with spam, attempts to steal from community funds and make the protocol worse, some proposals did pass. Let’s be real, most voters are uninformed and will follow the majority, that’s a dangerous thing to have on an open governance platform if the consequences for abuse aren’t harsh.

3

I can’t give you another specific protocol which has all of what I said, I’m gathering this from observations of other protocols.

But if you need me to list a few:

  1. Tokemak - Almost forces participation in return for rewards, force is a strong term, we would like to encourage, rather than force.
  2. Spectrum - up to 55% APR for long term governance staking.

I don’t need to make any specific mention about protocols that have multiple polls going on with sufficient community engagement.

But if you read closely, I did mention, in agreement with yourself that I wouldn’t have liked it to be 100 ANC.

However the cost to run a poll should not be that high. The major issue is the potential (and seemingly quite likely) loss of the 1000 ANC not because the poll was not agreed on, but because nobody participated.

I think above everything else, we should have a reject poll, and abstain poll options. Because these will help define participation. One should be able to join a poll and abstain because they agree with neither solution being proposed. Similarly, one should be able to voice that a poll shouldn’t be there through voting, to disincentive governance attacks. Subsequently we need to look at better governance rewards, because the opportunity cost of participating in governance on Anchor is prohibitively high.

4

Thank you @DaManiac - It seems that you have put a lot of thoughts into this post.

Here are some of my thoughts:

  1. Costs of starting a poll
    I see your point and definitely it sucks for the person proposing a poll if the quorum fails because of lack of participation.
    I am very conflicted about this, as having a high poll costs can combat spamming of proposal which we obviously do not want. 1.000 ANC does seem very high, however, if a “small fish” has a convincing proposal and puts it up to discussion in the forum. I am sure that a bigger fish would be happy to sponsor that poll. Maybe lowering it to 500 ANC could be a good start.

  2. Direct benefits
    Here I disagree. Poll creators shouldn’t directly monetarily benefit from creating a poll. By providing a good suggestion, the network should gain value (as it will become a better product) and therefore, the poll creator will indirectly from it.
    While 10% participation can seem high, I find it difficult to find a substition. Do we really want changes to be implemented that are only supported by e.g. 2% of the token holders?

  3. Multiple options
    Here I disagree too. I have my doubts that users are not participating because they are lacking options to vote. It’s simply that there are many users that are simply not participating in the protocol and its development. Users that are currently not voting are effectively abstaining from the poll. Additionally, it’s IMO very hard to differentiate between a simple NO and a REJECT. Each user is unique and therefore has different moral lines that they draw. What might just be a NO for me because I don’t like the specifics of the poll, could be a REJECT by you because you see it as an attack on the network (just an example).

  4. Participation rewards
    I think you meant, that ANC is inflationary :slight_smile: .

Maybe an idea to incentivize participation could be that poll participants get a higher ANC Governance reward compared to users that simply stake but do not participate?

This would encourage users to actively participate in polls as it gives them a monetary reward.

Another idea to increase user participation is to have a better UI in the poll field, that gives a quick overview what is being voted on and what are the pros & cons of it.

In this thread: UI Improval for Poll Voting some users have already discussed certain UI ideas that could help increase user participation.

It would be great, if this idea could get some more traction again (it got a bit lost during the LTV raise)

5

The Direct benefits only serve to offset the risk of failing to meet quorum. Nobody should be forced into a net negative chance situation especially when they are taking one for the team.

If you can neutralise the possible loss from a lack of quorum of being rejected, then there shouldn’t be any rewards. But as long as there is a possibility of a loss there should be a compensation for that.

Multiple options are never meant to encourage voting, they are there wen a scenario needs more definition. Like say for example, if we can agree that we need to extend LP rewards, but some people might find 6 months to short, so we can give 6, and 9 months as options, as well as a No, I don’t believe we need rewards, and then a possible, reject option, in the case of the something like the degenbox poll.

I did mention, that definitely we need higher rewards for participation. else, one might as well stay in the liquidity pool, or not own ANC outright.

As you pointed out, we can’t multitask, given the billions that are sitting in our vaults, people in general aren’t interested in making the protocol a better place.

6

@DaManiac this seems to tackle two ideas:

  1. Incentivized voting / lack of governance participation

  2. An adjusted minimum ANC deposit for proposals

I will comment on the latter first and push back on this line a bit:

I do think there is merit in requiring certain thresholds to post a proposal - it ensures more quality and more effort per proposal.

Terra recently dealt with a similar issue, resulting in lowering the minimum deposit and shorting the length of voting periods. I will use this screenshot as an example:

Screen Shot 2022-02-10 at 3.10.51 PM
Screen Shot 2022-02-10 at 3.10.51 PM2264×1318 245 KB

Notice the range of topics and support? I would argue these types of proposals (more current event focused, fewer product improvements) would not occur with a larger minimum.

As for the first point above, there are multiple ways to creatively think about incentivizing voting. Algorand is one example, rewarding “Governors” 15-18% APY for voting during each Epoch.

This discussion should be ongoing. I wonder if there is an opportunity to partner with tooling such as Snapshot to bring their technology to the Terra ecosystem?

1 Like
7

Thanks for pushing this, everyone. I agree that the gov side needs some serious work.

This was up for a vote and failed. It seemed there wasn’t enough consensus on the mechanism for this and levels before the poll went up.

Generally, a UST amount that sets the ANC needed solves problems of constantly changing market moves. Ideally, 250 - 500 UST should be the range looked at from the last talks.

I agree and this can certainly be worked into a new gov contract.

That also brings up different quorum levels needed.
Emergency voting as Anchor goes cross-chain will be needed to the cross-chain attack vectors that can be opened. A further measure is emergency powers (as seen in THORchain, Maker DAO, and other protocols) to freeze the protocol.

Borrowing from the lessons THORchain learned in the hack: Creating emergency freeze voting parameters to pause_chain, pause_trading, and pause withdrawals

Time variables :

pause_duration or pause_delay estimates the amount of time needed to implement risk_paramenters updates need.

Also, a possible vault_solvancy calculation which runs total protocol equity for calculation of pro-rata withdrawals that allow depositors to have first rights to withdrawal deposits.

An emergency voting module would need to be created similar to THORchain or MakerDAO. Possible safeguards of abuse include requiring a large amount (100,000 plus) of ANC staked and locked in the emergency staking vault module for a period of 90 days and at risk of slashing if the community votes deem an executive vote was abused in any way. The 100,000 can then be unfrozen and returned after the community regular quorum passes the measure.

An emergency freeze could be the preliminary action needed before a further emergency “risk” parameter can be changed. Exercising this function would allow executive modification on all Anchor “risk” parameters such as liquidation thresholds etc.

This also brings up quorum levels. At current, even if vote levels hit 30-50% percent approval, it still takes 3 days to execute the poll. In emergencies situations, we need a way to pass execute in less than 24 hours when there is unanimous consent in an emergency vote.

We need to change the tokenmomics. It’s been clear for too long with no action and that is inexcusable. I have been hearing whisperings from big stakeholders and new protocols that plan to be bigger stakeholders. More on that as I clean some up my research upon this topic.

Also on that note of making gov better, I have done some research on delegated voting. Its clear participation in gov here is horrible and some people still won’t vote no matter what changes are made - welcome to defi right now. One around this is delegated voting among uninformed voters who might have otherwise increased the freeloader problem or worse vote in an uninformed way. There are many passive stakeholders that might want to delegate voting power to DAO board members or prominent community members. It opens up the type of informed voting needed for a mature governance platform.

Creating delegate voting pools similar validators would allow stakers to either create a new delegation pool or delegate voting to a current pool. UI would have profiles and voting records perhaps powered by Anchorian score that measures number or props, successful votes etc.

Proxy voting would enable users to delegate their votes to vote for these elected delegates.

Agree that this would be good to work into the future new gov and token work coming.

1 Like
8

@fig what you raised is valid when viewed in isolation.

The idea of lowering voting cost / combining with direct rewards for sponsoring a poll, needs to be done in conjunction with the ability to shoot down bad poll ideas, to ensure these will not have their poll stakes refunded, through active voting rejection.

Think of this thought flow process.

  1. Should I risk my tokens for no net reward? If your answer is yes, then it ends here, no poll incentivisation. If the answer is no, then we clearly need some form of incentive.

  2. How do we prevent mass spamming of bad ideas attempting to game the poll incentivisation rewards. If we should do nothing, then clearly incentivisation will fail, and can can’t proceed, but if the answer is yes, we should active police against that, then what I suggested as part of a poll reject function is a possible solution and that should be counteract bad poll ideas.

I do not believe we should shorten voting periods, I believe we should increase the time frame. But alongside that we need to increase the voting incentive, to reduce the opportunity cost of the ANC tokens being locked up.

Imagine locking your ANC up in a bear market, and the price moves downwards 5% a day for 7 days. We don’t’ only want bull market voting. So we really need to increase the voting incentives.

@bitn8 I don’t believe delegated voting is in any way increasing community involvement, we want more people to click the button and be involved, not to hand over the responsibility to someone else.

If we incentivise voting financially, and possibly even allow voting with ANC-UST LP tokens, this might allow people to negate the opportunity cost of voting and bring more numbers. ANC-UST LP tokens would receive a smaller rewards though.

9

I don’t agree with giving incentives to vote, the whole idea is that we’re investors in the protocol and making it better is the incentive. People in crypto are way to reliant in immediate gratification, and that’s not a sustainable model, Mirror has proven that incentives will lead to abuse and poor governance.

I’ve said this in another post but the focus right now is fixing Anchor revenue model, and once that’s done holding ANC will become attractive because of the revenue share (right now via buyback, could be improved in the future). The incentives are just that, working on fixing Anchor so we can reach a sustainable model.

10

We already have incentives to vote, but the level of incentive isn’t sufficient to compete with the opportunity cost.

11

You have incentives to stake, and that’s due to the current revenue share model, you don’t need to vote.

12

Please stake more anc everyone will be happy then the apy and apr are fine