Done 
Hopefully we can get some more votes but it seems the majority are in favor of this 
1 Like
Love this! HarmonyHUB definitely supports!
2 Likes
Hello everyone,
I have some concerns about this proposal and will have to vote NO.
While I can see positive effect with reducing the period from 30 days to 7 epochs (I have been myself a victim of the spamming attack), I can also see the perverse effect for the small/unelected validators.
Giving them only a period of 7 epochs (5 days) instead of 30 epochs with a higher APY will reduce their attractiveness to gain new delegators.
Am I the only validator who reminds how many new delegators I had during my first 30 epochs once elected?
If I should vote for my own interest, I would vote yes on this.
If I should vote for the interests of the new validators and the decentralization of the protocol, I would vote NO.
The second option has the priority for me.
With HIP-14, the small/new validators will quickly disappear from the top of the list ordered by APY and the bigger validators will remain on top (because they can eat the keys), and theyāll continue to swallow the delegations.
If you make a new proposal just to fix the average APY (without changing the 30 epochs) to count also the unelected epochs, I might consider voting yes.
An other possibility might be to remove the APY from the staking portal (just like terra did) and only show a global APY (average of all validators displayed on top of the page). But if we do that we still need a new incentive to stake on smaller validators.
Thank you for attention!
2 Likes
Sorry but Iām not one to perpetuate a lie, because in essence that is exactly what allowing an APY to exist visually that isnāt factual is. Delegators deserve the truth period; not a benefit they think they are getting but in reality arenāt.
Being transparent is always the way to go, hiding truths from the community is not. Especially when the reasoning is to help validators out at the expense of hiding facts from the delegators.
1 Like
Their high APY is not a lie, thatās because they were elected with a small stake.
What is a ālieā is that we donāt count the unelected epochs in the average, but when I read your specification there is nothing mentioned about that.
Your specs :
- replace 30 epochs average by a 7epochs average
- remove or disclaimer for ER info in smartstake
So instead of a 30 epochs ālieā you propose a 7 epochs ālieā.
2 Likes
You either donāt get it or refuse to admit what Iām stating. The APY is a ālieā if itās not what the delegators are getting. Regardless of why it is shown, the delegators deserve a more accurate representation of what they can expect to earn. Period.
Smartstake has already been updated to reflect that the ER stats arenāt accurate.
A 7 epoch average is much more of a factual representation than 30 epoch average.
Iām not sure what you want with unelected epochs being counted. If they WERE counted then they would have a drastic effect on the average as they would have a 0% assigned. If anything NOT counting unelected epochs is beneficial to those validatorsā¦
1 Like
Well, if a validator missed an epoch, this should reflect into his average return, isnāt it? Sounds more logical and transparent like you want.
Maybe only start to display the average once a validator has been elected at least once, this will prevent to show 0%.
And your argument saying that with only 7 epochs the validator will be affected for a shorter period of time in case of problem is true BUT it will affect you way harder.
1 bad epoch in a 7 epoch average will affect your stats a lot with all the consequences we know in term of potential undelegation
1 Like
I tend to agree with @ONE4All here. I think 7 epochs is too short of an average considering shards rebalance each epoch and one bad epoch due to maintenance or other reasons outside of a validatorās control could severely impact validators sorting by ER and uptime, which obviously wouldnāt be indicative of their overall performance. A 30 epoch average gives a delegator a more comprehensive view of a validatorās performance. I would like to see a more interactive portal where delegators can select their own time frames or have returns from multiple periods presented (last epoch, 7 epoch, 30 epoch, 90 epoch, lifetime etc), which would provide the maximum benefit to delegators.
1 Like
The discussion is great and thank you guys for the different perspectives. I would like to point out that this has been on the talk forum for 4 months. Waiting until itās being voted on (due to overwhelming 96% approval on the poll) to provide feedback isnāt the intention of the talk forum. The intention is to discuss the proposal, change it for the better if need be, and then put it up for a Snapshot vote to be ratified. Or kill the proposal here so as not to waste peopleās time with voting. I feel like we could have came to a consensus on the best product to put forward to ensure a great chance at passing; now it appears to be a potential proposal that is in jeopardy of not passing even though it had a 96% approval out of 67 votes.
1 Like
When we show the āinformation statementā, would this include also the unstaking period?
I donāt think this is communitcated enough in the staking dashboard and upsets a lot of delegators.
2 Likes
Absolutely, the 7 epoch un-staking time period should be included there.
1 Like
I can see both sides. I think what this gets down to is different takes on the following:
The ER displayed upon clicking āValidatorsā on the staking portal is a 30 day average and often not a factual display of current expected return.
This is both true and untrue.
Why it's true
It's true in that a delegator cannot expect a return anywhere near the return of a validator's first election. To insinuate that they can is a "lie".Why it's untrue
During network distress (spam attack, etc), some validators had better infra / disaster recovery / worked tirelessly / emptied their pockets to keep signing for their delegators. Some did not. To insinuate to delegators that the past 7 days represents validators' reliability / ER when things go south is a "lie".Suggestion
What if we did something like this?
That way, visibility is given to āstorms weatheredā as well as āwhatās todayās rateā - not unlike analysis in legacy finance.
Closing thoughts
I think I'm a "no" for any changes that don't show ER at different points in time like one might expect when evaluating fund managers or something.The entire experiment of decentralized decision-making, DAOs, and ONE rides on the proposition that we can swap ideas, not only when things are unanimous, but when there are differing opinions. Kudos to the community for bravely traveling into the unknown with me in good faith.... :grinning:
This passed the Snapshot vote over 2 weeks ago and was submitted as a github request for implementation.
2 Likes