Should wallets penalize pools in the ranking and how could that look like?

Staking & Delegation Stake Delegation
,
1

Hello, based on this tweet from Nico… https://twitter.com/NicoArqueros/status/1321159036675461125

…I would like to take the opportunity to bring this discussion into the forum.

There are two points in the Tweet to steer the discussion:
First, the penalization of pools that are below a certain fee/margin.

Second, whether and to what extent a wallet could penalize multiple pools in ranking and how that could look like.

Since I have already read a few tweets about this I would like to say as a preface that a certain wallet ranking was never in the scope of the Cardano protocol but there is a specification for the ranking in this document, which the wallets can (or cannot) adhere to https://hydra.iohk.io/build/790053/download/1/delegation_design_spec.pdf (from page 40 and following)

2 Likes
2

Daedalus currently has incentive for you to chose 0%, changing that would be a good start.

5 Likes
3

Assuming the protocol is designed to be fair, why is there any ranking system at all. (#1,#2etc)

In my limited understanding of the protocol:
Rewards should be random and designed so a pools delegators are getting equal returns on average long term: Around 5.6%.
Making a ranking system a strange concept.

If that was the case then the only true way of differentiating/ranking would be how often the pool is online. (And fee differences)

I learnt recently that The 340ada fixed fee is a big problem for new entrants. So that is something broken in the protocol.

Censoring pools is not a solution and Removing all 0% fee pools from your wallet software will probably lose you users if delegators learn and realise.

P.S. wasn’t Yoroi development paid for by the Cardano community so it concerns me that Emurgo or anyone from Emurgo can so easily change it to do what they arbitrarily want.

1 Like
4

An impartial wallet should only present randomly the information giving the opportunity to delegator to filter and sort by differtent parameters. In this way we present the information to the delegator and delegator is what finally make a decision acording his/her investigation. It should be like in the stock market, stocks are not ranked at all, is the researching made what make a people to choose one or another.

That can lead people to get acquainted with delegations flags, parameters and metrics and not only delegate blindly with popular pools, retired or saturated pools.

But penalizing to me it’s definitely not the concept of freedom.

2 Likes
5

Having a list from 1 - (# of stake pools) is too granular and misleading. There might be no functional difference between #1 and #15, but the rankings imply otherwise.
There should be 3 groups - Green, Yellow, and Red. The order within the groups should be random, with a randomizer button and the ability to filter based on many parameters to find a pool that fits the user.

Green indicates pools that have minted >Z% of their slot leader positions over the last X epochs. Online and making blocks on time is THE criteria for a good pool. These pools should have a minimum pledge value.

Yellow have minted <Z% of their slot leader positions over the last X epochs, are >Y% saturation, or have little to no pledge.

Red are pools that people should be wary of delegating to. These are pools that are saturated, retiring, regularly miss blocks, have a higher than W% variable fee, have a higher than V epoch fee, or etc.

As long as pledge has no value there will be no way to stop people from making multiple pools. The incentive needs to be driven by greed - make more money with more pledge.

3 Likes
6

Thanks for bringing up the subject. I understand that most people engage in these discussions with the best intentions but to be honest, ranking is pretty subjective and adding the 340 ADA to come up with a real % thwarts most chances that the average stake holder will delegate to a smaller pool.
Then you have a lot of the community influencers stating that small pools are not sustainable because they choose to lower their cost in order to be competitive (it’s pretty much the only edge we have). We have to think beyond first world countries, I can think of several examples where even minting a single block per epoch would pay for the pool infrastructure and allow the SPO to actually live off their investment.

4 Likes
7

Thanks avilsmeier,

since I’ve been a part of the community here, I’ve found these discussions odd. I’ve tried to pay close attention when people tried to explain why 0% was ‘wrong’, but it’s never made sense to me.

How is it that it is impossible for people being able to imagine someone running a stake pool for the fun of it? for the sake of contributing something to a larger purpose than making money.

I’ve come to believe it is because of the basic premise of game theory that everyone is acting in their own interests all of the time.

It’s like, can’t we work with game theory as far as it goes, and then accept that altruism also exists?

My 2 lovelaces…

1 Like
8

Personally I have stated this several times and will do so again. From the protocol and spec perspectives the ranking pertaining to Daedalus is more or less functioning as they intend however this is only good from a perspective on who is winning the defined game at hand ( stake pool operator view ).

These wallets are the Highways to the network as such they are the main source of information between investor (delegate/stakeholder) and pool as such they are very much in control of what makes or breaks a pool or pool operator, in my opinion the current system is broken as it has little to no solid information for the delegate/stakeholders to make informed decisions as it does not take the “investor” view point into account and it should. Meaning if you were to open the wallet as an investor and look to the pool with the highest return on your staking investment you would not be able to make that choice clearly from the information given and this should be taken into account as the delegate/stakeholders are key in this system.

Now this is all well and good for the start of the argument just this varying viewpoint however, There is also the argument on the UI/layout of the wallets as the network grows these cannot possibly scale and be useful as users will not bother to sift 1-1200+ tiles or info lines on pools with little to no performance difference between them. More and more this shows to be true currently given time the protocol punishes and rewards pools equally as such returns even out over time as intended.

A suggestion on this could be include a “staking” (equivalent to "I’m feeling lucky on google search) button that functions to be a wallets algorithmic choice between performant pools on an agreed fair ranking system pertaining to data points that rotates a selection of pools. When these pools reach a level of sustainability or have undesirable performance results they are moved from the rotating list for auto delegation. This would not be the sole way to select pools as you would still be able to select your favorite pools like current. By no means is this a best option but it is a start in the discussion on a path to a better solution.

On the subject of current ranking and fees, removing the fixed 340 is not the solution this is a common ground and a starting point this should not be removed for convenience just like pledge should not be gamed as a cost to security of the network. To remove the fixed 340 you just open the door for losses to continue as pools race all the way to 0 fee and 0% this is idiotic and would just hurt the network even more. Where the fixed cost has conflict currently is in the “Adapools” ranking system used directly in Yoroi-wallet when they impose a “Cost” line in the ranking that skews the results in favor of large pools over small, example: my pool with a fixed 340 +2% margin reflected a 18%-20% cost to stakeholders or more. How do you explain that as fair? I personally have yet to see how they calculate this “Cost” total (or many of their unique ranking points) but on others research and my own when you compare the returns for a wallet holding the same amount on a large pool or a small pool these “costs” have not been reflected and hurt small or starting pools just by being included in the wallet they directly contradict the “30 day ROA” You can’t have a competitive ROA and a high “cost” to users at the same time.

These are just a few of my thoughts on the subjects and not a magic bullet as there is none for a instant solution. We have to come to a consensus as a community to correct these and other issues as they arise if we plan to grow this network to the full potential so far this is lacking.

For those of you on the “I want to charge 0% and 0 fixed fee” be realistic this is a gimmick at best. If you are running the core of the network you are not approaching it as a serious business and you are subsidizing your running costs as you assume this will give you the advantage in the race to the bottom as a marketing attempt, even charities have running costs. If you are doing it for the benefit of the network and for the cause then you would truly understand this hurts the network and not question the fee floor in the first place as not everyone can run this way and its unhealthy to the network and was never the intention look to IOG pools if you need confirmation on this.

I ask you to look at the other networks offering POS and their average fees, running costs and costs for entry into the systems even if their offerings are worse, you will find they are substantially higher on products that are arguably inferior to the scale and scope of this project.

Yes i know “anybody can run a pool i do what i want” but that does not mean they should be running a pool nor that the pool will be profitable or that it is healthy for the network.

For some reason day one on ITN operators looked at 1% and arrived at the conclusion low fee was the best idea to gain delegated stake but they did not understand his intentions on running multiple pools (this makes up for the low fee and multiplies the return) or have the intentions on running several pools to compete to his level to this day they still don’t grasp the concept as such the network raced to ultra low fees and we arrived to where we are today unsustainable at scale because the network is not supporting all the pools and we have barely scratched the surface of what a stake pool is or does on the network and what it will take to run them in the future.

Yes i went a little off the rails here but we need to look more into this beyond a “haves and have not” argument. The network can support the amount of pools we have here already had stake dispersed evenly from the start with a more realistic saturation level and this would not be the discussion as all pools would equally share the load. The doing nothing to this point to correct this issue or openly addressing it with a solution and a timeline only magnifies it.

4 Likes
9

If a ‘small’ pool mines 1 block per epoch average, and gets rewards of 1500ada for that block, then the 340ada fixed fee FORCES that small pool to have 25% approx fees.
Thanks to TOPO stake pool for explaining this concept to me in telegram.

I’m not advocating 0% fees.

But the bigger problem is the fixed 340ada. It seems like a fail in the protocol to me, which discourages a rational delegator from delegating to small pools (because they won’t be able to get their ROS attractive due to the 25% fee)

It seems it would be better to have a minimum 1% or 2% etc variable fee than it would to have the 340ada fixed fee. With the assumption we want the protocol to be fair and most decentralised.

10

Small stake pools are the same problem as multiple stake pools. The 1PC conglomerate is a group of small stake pools. This is a PROOF OF STAKE coin. It is silly to give people with small stake (small stake pools) control over the network.

11

That’s not the case in practice for all “small” pools. You can see this if you compare like addresses from a large pool to a smaller pool. The rewards are proportional to stake always, where this changes is in the luck of a pool if a pool was projected based on its stake to only produce .5 blocks an epoch on average and it produced 3 blocks the rewards are boosted. This still does work to average the return out over time between large and small pools the difference is how many delegates are in the pool and how the “fixed” fee is split among those delegated to the pool and the fee charged. The way Adapools is reflecting this or calculating it does not tell the whole story. I know this from first hand experience and the returns from my pool vs that of a large pool. Yes we are running at 0% +340 fixed to offset any assumed loss in returns to our stakeholders (this doesn’t mean the operation is smaller or costs less to run) because we see a lower consistent block return however averaged out from the start so far this has kept the returns the same or higher to our stakeholders you can look at the data provided by pooltool to verify this (use similar addresses totals to compare over the same time). At the same time Adapools shows us for an average cost of over 17% in their cost column which is wrong as our delegates are not losing 17% of their returns nor are we losing that for them. The fixed fee is to put everybody at the same starting point and yes it levies a higher tax on a pool with fewer stakeholders (by numbers as less stakeholders share the fee) but just a variable fee alone does not make up the running costs for a small pool or gain it guaranteed stake and doing away with it just allows the race to 0 to continue. There has to be a middle ground and a floor set to keep the network viable for the future and attractive for professional operations as they are vital to the core of the network. There is a balance to maintain just like the protocol is trying to balance for the rewards.

1 Like
12

I appreciate so much of what you say, but this statement above, i do not get.

What is unhealthy for the network? To me, what seems unhealthy for the network when it comes to stake pool operators is when they get picked to process transactions and they fail.

Isn’t that the bottom line? Do they produce blocks when they get the chance or not? I would think that a stake pool that fails to produce the blocks when they get the chance will indeed fall by the wayside as time goes on. Isn’t this true?

I feel like this argumentation for pool operators to charge a high enough fee is based on some level of fear of being ‘wal-marted’ out of the ecosystem.

As an investor, hoping for a return on my investment, I picked a pool based on some comments here in the forum that I saw one operator make. Seemed like a nice guy, made a professional impression in all respects. But he didn’t produce blocks. I have no idea why. It wasn’t because of his pledge, it was close to what a lot of them put up, he charged 1%, so he’s not trying to game anything in my view.

So I switched pools, went to the number 1 ranking the day I switched, and they crank out the blocks, enough that I’m earning what I was led to believe.

Now, if I did indeed have a true friend operating a pool for reasons of a desire to be a part of the Cardano ecosystem, I would probably delegate to this friend. I wouldn’t particularly care what he charged if I felt his intentions were in alignment with the ecosystem, and if he wasn’t being picked, like 1050 are not being picked every epoch, I would hang with them until K was increased.

But I don’t know anything about these people or about you Glitch. I have been hoping for Cardano, and for the 1050 stake pool operators that in every epoch are not chosen, that all of the pools are competent, and have the staying power so that eventually as Cardano grows, all pools are required to process transactions, all 1200+.

But for me to continue to follow these conversations, somebody please tell me why someone working for less return than ‘most’ is bad for the network.

Can you do that Glitch? Please? I want to believe you, and others, are writing so much for a good reason. But I don’t feel like I’ve heard it yet.

Peace

1 Like
13

I touched on this in my post.

There are associated costs with running dedicated stake pools full time It’s not realistic to expect the core of the network will run on the base 340 fee and provides a level of uptime low latency and performance day in and out to maintain the network at the core.

This was never the intention or design, why would it reward the delegates and stakeholders but not the SPO/owners putting in the time, effort, money, manpower to maintain and run the network? Answer: it didn’t, This all stems from “FOMO” of operators that started on ITN and carried its way into Main net from the extrmely small K setting at the launch. Why was / is it FOMO? because this turned into an advertisement within the wallets to gain stake holders then raise fees later (low hanging fruit). This is purely an attempt to outlast others and gain the most delegates and raise the fees later after others give up in hopes that the base value increases to recover any losses.

What happens if/when the value does not raise and these pools with low fees and are full of delegates shutdown because they can no longer cover the running costs, and the average fees remain at this low rate or 0? Does it attract other operators to the network? Not really so the larger operations get more delegation and keep expanding centralizing the network more and more to mimic BTC the opposite of what we want.

With these reduced / minimal fees unless an operator is subsidizing his operation on the back of other servers (less than ideal) they have control of or have and an unusually large pledge fund they will be cutting corners to reduce running costs. This could likely lead to hosting several pools on the same central servers or relays further centralizing the network for a large core of the network. The larger they grow and consolidate stake makes the network more susceptible to attack or interruption if there is an outage, as there ends up being multiple pools located on/in the same location which risks a large portion of the network in relation to the size of those pools.

The other direction this takes would be pools being setup on home connections scattered around the globe and as nice as this sounds this will never allow the network to scale as it is intended as home connections are less reliable. The core part of this network needs to be on server grade hardware with direct connection to the core of the internet to function with the speed and performance a network such as this requires.

Bottom line here its bad for the network to run on fumes and expect exceptional performance as an always on global infrastructure touted as being the future, then expect it to run on hopes and dreams while there are real costs and investments in play here much like any business.

1 Like
14

Well put.
If small stake pools are economically unsustainable then most will either increase stake or disappear over time.

If 0%+340 pools are unsustainable they will disappear over time.

If ranking makes a difference for pools then the system will be gamed.

A rational delegator will choose a pool with nonzero but less than full saturation and look for low fees only within that group.

1 Like
15

I agree with you waldmops.

This has been my feeling all along, if their setup fails, they will disappear, if they do the job when they get the chance, that is the proof of who and what they are.

I don’t see how a small pool affects or controls anything other than their own lives.

The opportunity to do it, if competent, is kind of what is promised by engineering the blockchain so that affordable hardware and energy consumption are enough to be a part of it. At least I thought so.

I do understand @glitch_04 and his comments about the need for a hard pipeline to the internet. But some people have that. And if I did, i might have seriously considered starting a pool.

Thanks everyone for the conversation.

16

Reading such academic paper backing up the idea adopted for managing pools is refreshing. That is why I like Cardano. Everything they do, there is credible research behind it.

17

Pool performance is one criterion for ranking, and not all pools will have the same performance due to their ability to maintain their pool uptime. If their servers are down, they won’t make any blocks.