Summary: reflections of rewards, pools

If the intention is that there are between 1000 and 2000 pools. The pools of 10 million upwards should have the same probability of being chosen. So if there are 20 billion staked we would have 2000 pools with the same incentives. In addition, there should be a maximum amount of coins established per pool and some minimum and maximum cost scales. From 3% to 10%? In the current testnet way, after delegating in several pools I have realized that to receive coins, I need to delegate in a pool of at least 3 million- In addition, it costs nothing at the time because I It looks like a scam and if all the pools charged, they would discourage staking. Now, all pools with less than 4 million do not get even 1 coin. In the end there will be 100 pools with 100 million each or more and then 1000 stakes will not be achieved.
I have decided to ensure less rewards by delegating in groups of 40 million with a 5% rate, because if the rate is too cheap, the pool grows too much and the rewards for everyone fall.

1 Like

yes… the pools are too centralised

maybe IOHK were expecting ~100 pools in the testnet which would suit the current saturation level but clearly there are far more pools than anticipated

either the saturation point is increased to some where in the range of 200 to 400 to take advantage of the number of staking pools or 300+ staking pools disappear

if the desire is to decentralise the network as much as possible then the saturation point / protocols K value clearly needs to be adjusted to take full advantage of decentralisation

2 Likes

Part of the problem might be that one has to set up a separate Reward Wallet for every pool one wants to stake to which is cumbersome and is perhaps limiting how many pools get stakes.

At this point, with sub 100k ADA to play with on the test net, if I could delegate to as many pools as I liked without setting up a dedicated wallet I would spread the stake out to more pools just to help get everyone up and running.

1 Like

Yep, we have been active on social media advertising the pool, helping people out on telegram, announced a competition to return 100% of rewards to delegators, even considered buying test ADA but we still can’t get any stake.

Our fixed tax is 0 with a pool operator reward of 2%. While I couldn’t justify a higher tax (unless we were going to employ a developer to start building tools), we can’t really afford to reduce it any further while still being able to expand the way we would like to scale our infrastructure.

At this rate, all we can do is hope the incentives game theory works as planned before the saturated pools start creating pool number 2, 3, 4…

Coupled with this is the poor performance algorithm in the Delegation Centre. We have been ranked 3rd from last the entire time since the first epoch other than very brief visits to a higher ranking when we do manage to beat the odds and produce a block. We report our tip to pooltool.io (consistently green), we have produced every single one of the 3 blocks we have been elected to create and our performance metric is still 0%.

Additionally, we are ranked below duplicate abandoned pools that have been decommissioned and are inactive.

I’ve been a member of the community for years and it’s difficult to watch people spinning up pools with no real understanding or belief in the project, simply to profit, and then get assigned a huge stake from IOHK while our efforts seem unviable.

We have every intention of sticking it out and continuing to work hard, but we really feel that saturation should drop to 0.1% - 0.5% and hope CF, Emurgo and IOHK will consider delegating to some of the smaller pools.

Web - https://cardanopool.party
Ticker - PARTY
Telegram - https://t.me/Cardano_Pool_Party
Twitter - https://twitter.com/ADAPoolParty

7 Likes

Hi Matthijs,

that’s really sad to hear. I think that, from the beginning, the staking did not work as expected: Pools with 6% share created 11% of the blocks while many small pools didn’t get the chance to produce one single block. If there is a random number generator behind this, it should be checked and corrected. In the meantime this looks better but ZZZ still creates twice as much blocks as they should concerning their percentage of the overall stake.

I don’t think that reducing your pool operator rewards makes it any better. The problem is that the standard user delegates to the uppermost pool in the ranking of Daedalus and this is not necessaryly the best pool for him. The Daedalus ranking must take the saturation into consideration, otherwise there should be better no ranking at all.

By the way: As I wrote here I think they reduced the saturation level to 0.5% of the total stake which would allow more small pools to grow. If someone could confirm or negate, this would be fine.

2 Likes

Hey Matthijis, nice webpage. If staking were easier as I described above I would think more people would be willing to spread their stake around to help pools in your situation.

I’m guessing these are the problems the testnet was designed to identify and I’m pretty confident IOHK et al with sort them out.

2 Likes

I’m not a math / game theory wiz but what if a 1% to 5% of every blocks was awarded randomly? Is there a number that would keep small pools in the game?

I don’t like that idea. The reward is not a gift but a, hmmm, “reward” for a delivered work, namely be ready and create a block when chosen. If the act of chosing is really random, then every pool has a chance, even the ones with very few stake. Maybe they will get a block only every 10 epochs? But then the reward will be so much for this single block that they can sit and wait again for some epochs. In large pools the situation is less stochastic. Maybe they create 35 blocks one epoch and 40 the other epoch. At least they can be quite sure that there is some kind of reward each time but this has to be divided by many delegators according to their stake.

I have the feeling that the problem sits somewhere else: A pool which wasn’t chosen to produce a block up to now gets a performance of 0% which, as far as I understand it, reduces it’s probability to be chosen for producing a block. This makes it much harder for new pools to come up. If this is really the problem, then maybe it could be solved easily by setting “virgin” pools to 50% performance instead? But I’m not really deep into this, I’m just interested, read what I can find and play around with Daedalus to learn how the diva works.

To be clear I mean what if the chance to create the block was awarded randomly every 1 out of 100 blocks. So that someone starting out with a zero stake has a better chance to create a block and then get the reward.

I found a great thread on Reddit about how to chose a stake pool in the testnet.

1 Like

Goes against the most fundamental principle of POS.

1 Like

Agreed. Just talking about getting things started so small pools have a chance to get thier feet wet.

Hi @matthijs.aeon, Yes it’s frustrating, but unfortunately that’s how it will and should work. The k should be 1000 and what you refer to saturation point is a0 which should be decided for the mainnet (based on the big ITN experiment, I think) and can have a value between 0..1, means 0%…100% of of the 1/k of the available stakes (as I remember, and not the total stakes i.e. 31bn initially in the mainnet), means your % of your expected/idealistic pool’s stake.

So a0=0.1 means that 10% of the 1/k * S, (not really as it is simplified a bit for digesting easy), and I think that’s a fair value of it to prevent Sybill Attack (by creating as many pool as we like, attract the delegators and attack the system /w their stake.

E.g. I could create 500 pools from almost no money (not really), then attract them, by any marketing tools and then bring the network down if we reach 51% or change to future/near past history, making me the richest in the world.:). I would say 0.1 should be feasible, and yes that means that my pool will be ruled out in mainnet as it would need around 3.1mn for the 31bn all stakes. But, there are some options:

  • Make a group of trusted honest ppl and pledge all yours money together into a bucket. As a pool (with one operator) can have multiple owners. It can be achieved /w ring/group signatures (relative:) easily.
  • Risk it with lower pledge and do PR for attracting delegators.
  • Create a really transparent Stake Pool Alliance from pools only with the very trust worthy, with long history in the Cardano community, who are transparent and decentralisation focused, and they can make a pledge to have only 1 pool operated by each of them. So, that could be a core of the decentralisation, but of course they will be constantly attacked, criticised to bend the public’s opinion, as realistically we are talking abt money and a lot of money in the future. And this will attract the psycho and sociopath like poo the flies (“Biiiitcoooooooneeeeeeect”:).

So, the registration fee, the pledge and all other things is all about to secure the network from the weakest links of the chain us the humans.
And, I am very satisfied /w the current state of the shelley blockhain, as we could not bring it down with almost 6 times more number of pools (unconfigured, misbehaving, unreachable etc) than the current k (100).

Anyway, it was long, but I think my pool will be long forgotten in 2-3yrs from now.
Check the digestible details here.

6 Likes

Pls give it a second thoughts.

Are you sure about this? I think, k right now is more like 200. Why? When I look into adapools.org, right now the maximum total rewards (stakers + pool operator rewards) is 30.1k ADA. If I look at some smaller pools, I can find e.g. TIM with 0.39% stake and total rewards of 25.0k ADA. This shouldn’t be possible with k=100 as far as I understand it.

Do you know of any place where the actual value of k can be observed?

@_ilap Yor explaination above makes a lot of sense to me. Thanks