I think the current state is highly untruthful to the delegators and I doubt, that there is a high demand on what is my validator’s ER, but the interest is (understandably) more what is my ER when I choose this validator.
If there is high demand for that, we could add a column “ER before fee”. But as described above, I don’t think people care about it and if they do, they would go to smartstake to get more insights.
I agree, that it’s somewhat counterintuitive that commission can change at any point in time. But when showing the ‘fixed ER’ it would become more apparent in the dashboard, when the ER drops.
Showing a 30 epoch historical rate when the next epochs rate could be 0% or unelected is misleading as-is. Is there a strong push from the community for this change? It’s never been brought to our attention from any delegator. If anything, we’ve educated them that our ER doesn’t include the fee and delegators have become comfortable calculating the rewards themselves.
If the goal is to increase staking participation to new valdiators with less than 5% fee, I can definitely understand the motive then. Im just not sure if we should let fee weigh so heavily on the dashboard where a delegator is basically going with a validator just because their low fee improved their ER. The ER before fee column would be good so we’d essentially be adding a “Net Earnings” column so it shows gross earnings, fee, and net earnings.
The other thing to consider is if there’s a way to use historical commission rate and make it clear that the “expected return” column is actually “historical return based on 7 epoch average”.
HIP 14 was to change to 7 epoch average and remove latest expected return from the validator profile. It didn’t change the Expected Return column on the validator list. We shouldn’t claim an Expected Return for the following epoch since the validator could lose election or experience downtime, plus variable block reward rates and shard saturation will impact actual returns. The portal should instead say “Historical Return” with an asterisk or footnote denoting its based on a 7 epoch average.
My main point is the ER % is based on historical data whereas the portal is making it appear that’s the current ER % based on current data, which is what’s really misleading.
I agree. I think it should be changed. I assumed it included the validator list page. That’s my own fault. It would’ve been a good point to mention in the hip14 discussion thread
Although that still needs to be addressed, it shouldn’t prohibit the implementation of hip 23. the displayed “return” can deduct the commission - regardless of whether it’s historical or expected
If the idea is to help the new elected for the 100 first epochs by showing them a better APY, I can see the good intention.
But, in the long run, I am afraid this proposal will create an unhealthy competition between the validators to run at lowest fees possible to show a better ER.
Keep in mind that at this “game”, the largest validators will always have an advantage over the smaller ones.
Why ? A 10 keys validator can basically use the same server infrastructure than a single key validator, but with servers cost divided by 10.
I don’t know what is the best solution, maybe just removing the ER would be better. That’s maybe why Terra did that.
For these reasons, I cannot vote yes for this proposal.
I think this would probably have the side effect of directing more delegations to newer validators with 0% fees, but it isn’t my primary reason for supporting it. If I’m a delegator, I don’t care about the total rewards that the validator is generating - I care about what reward % I can expect to get from that validator. Since fees are taken out before rewards are distributed to delegators, I don’t think the pre-commission rate is as relevant to them.
@ericAZUR tested some people on twitter, and I did a similar exercise in other places to see if delegators knew how the fees worked. A lot of people thought that if the reward rate was 10% and the fees were 5% that the validator would only provide an APY of 5%. Some people also thought that the expected returns on the staking dashboard already subtracted out the fees, and I understand why they would think that. I think showing the gross APY, commission, and net APY together makes sense, but there are pretty good arguments for and against showing either of them alone.
I don’t think of this proposal as a proposal to support smaller validators. But more of being honest to our delegators. If you have an ER of 9% with a validator that has 100% commission, then your return is 0, not 9%.
There will always be those who want to understand the full details and do all the math. But a sizable group just wants to look at the highest interest and go there. I’ve seen this on multiple blockchains. All Cardano 3rd party analysis sites do it in the way proposed, FYI, and even still people make moves that don’t make sense.
It isn’t apy if it isn’t in your wallet. Everyone, whether they do the math themselves or rely on others just want to know what they are going to earn.
The problem is there’s no way to tell what a delegator is going to earn in the following epoch. All you can show is historical returns, and current commission rate multiplied by historical gross return isn’t a good indicator of expected return in my opinion.
Do you think that what we have today is a better indication of what they might expect in returns? I don’t think either option can be guaranteed to be extremely accurate, but I think the proposed option would be more correct most of the time.
Ideally we give delegators the facts: 1) historical gross return % and 2) current commission %. What is the point in calculating an inaccurate number and labeling it “expected return”? If anything. It creates potential liability for validators because it gives off the perception we’re guaranteeing returns when in reality our return could be 0% in the following epoch.
The point I was trying to make is that what we’re currently showing is even more inaccurate and still also called expected return. I agree that there’s no guarantee about whether that’s what the delegator will receive, but that’s currently the case as well.
Perhaps we should use historical data to calculate whether subtracting out commission or using the current method would be more accurate for what delegators actually received each epoch. If someone were to go through that exercise and it show that it’s generally more accurate to subtract out the commission, would you consider supporting this proposal?
From a delegator’s perspective I think this proposal is positive. It’s a more accurate representation of what the ER will be and that’s something that we should always strive for; transparency. However, the delegator could still get confused and think that this new ER is not including the commission. So, we either get rid of the commission column or add a new column for “ER before fee”.
From a small validator’s perspective, I think this proposal could threaten elected validators that only have 1 key and whose 100 epochs at 0% fee are over.
Validators with 0% will always be more competitive than small validators and there’s not much we will be able to do about it.
Validators with more keys, though, will be encouraged to aim for the lower bound to remain competitive against 0% fee validators.
Validators like me will be left on a limbo with less tools at our disposal to remain competitive (fee, keys).
This situation is already a reality, but this proposal might accentuate these issues. I’m inclined to support this proposal, but please address my concerns in case they aren’t accurate or valid.
The “expected return” will always be an estimation, that’s why it states ‘expected’. But providing an easy-to-understand metric that provides at least some insights is better than providing none. Also as a “newcomer”, I feel like @ericAZUR: Everywhere it is calculated that way. With your 2 columns, there is no column “my highest return”. And that’s what people are interested in.
I was sincerely taken into account, what if we would take the actual return after commission (RAC) for the last epochs and show the average, but it would be even worse.
I take the extreme example here:
If a validator had set his return to 0% and the RAC was 10%, now he increases it to 100%. Then it would still show some decreasing expected return for the following x epochs when there is no return to be expected.
@Rutilant_Hub in my eyes it’s not the problem that it’s accentuated, but that a proposal was passed, that gave smaller validators a small advantage with the 0% fee policy. But in reality, it’s not really visible for the common delegators: E.g. a validator with 10% fees and the same ER seems like the validator is just doing a better job.
I would like to leave this proposal simple and as is! I’m glad that we have a discussion about it, but I think it’s a very natural thing to do. As everywhere it’s calculated that way.
Whether a column “ER after commission” should be introduced would be a different topic.
I feel like I’m talking In circles here but there is zero logic in taking historical returns from the previous epochs, applying current commission rate, and labeling it “expected return” for the following epoch. It’s flat out wrong. There are so many variables that impact a validators return that it’s nearly impossible to estimate an expectation of return with any accuracy.
If we want to show historical returns (whether gross or net historical fees) and show the current commission rate, that is a much more transparent measure for a delegator to evaluate than an off-chain calculated return that holds no value in estimating actual returns going forward.