Staking Phase 4 - MIP-8

[schimih]

Introduction

Staking Phase 4.0, part of the Vega Upgrade (v1.7.0), marks an important enhancement to the validator mechanism in MultiversX Network, building on the foundation laid by the Sirius Upgrade (v1.6.0). Staking Phase 4.0 will contribute to the improvement of the network security, efficiency and decentralization. It will introduce key modifications to the staking process, particularly in the methods of selecting and managing validators that will participate in consensus (part of the active set of validators).

Improvements Down The Road

1. Soft Auction Mechanism vs. Fixed Staking

• Current: Validators need a minimum stake of 2.5k EGLD and they have to wait in the queue until someone frees a position in the active set of validators.
• v4 Enhancement: A dynamic auction system allows validators to bid for positions, creating a market-driven entry mechanism.

2. Node Selection Algorithm: Static vs. Dynamic

• Current: selection is fixed, when a node is jailed or unstaked, a new node from top of the queue is selected.
• v4 Enhancement: Incorporates a comprehensive algorithm considering top-up value for node selection.

3. Validator Management: Rigid vs. Adaptive

• Current: Limited flexibility in managing the entry and exit of validators.
• v4 Enhancement: Introduces automated, real-time management of validators based on continuous performance and network health metrics.

4. Network Scalability and Security

• Current: Limited scalability due to fixed node cap and potential security risks from centralization.
• v4 Enhancement: Increased network scalability and improved security through a diversified and dynamic validator base.

How will the upgrade happen?

A phased rollout is planned to ensure a smooth transition, each phase introducing key elements of the new system.
Let’s consider Epoch n - activation epoch of the upgrade.

Phase 1 → Epoch n: Introduction of Auction List and Removal of Staking Queue

• Objective: To transition from a static staking queue to a dynamic auction-based system.

• Implementation: Replace the fixed staking queue with an auction list where validators can bid for positions.
  

  

• NOTE:
  During the test phase on Testnet, we encountered an exploitable loophole regarding the maximum cap on nodes per Staking Provider (SP). Specifically, an SP could circumvent the cap by queuing additional nodes right before the activation of Phase 1 in the staking v4 mechanism. This tactic allowed SPs, regardless of their current node count, to potentially start staking v4 with an excessive number of nodes by leveraging any available eGLD, thus bypassing established rules.
  To address this issue and ensure fair play, an update ( PR #6042 ) on the MultiversX GitHub repository, Vega rc/v1.7.0 branch has been implemented. This update introduces a mechanism that automatically unstakes all nodes in the queue list at the onset of Phase 1 of staking v4 activation. Consequently, every SP will commence staking v4 with only their eligible nodes, adhering to the prescribed limit. This change safeguards the integrity of the staking process, ensuring compliance with established node limits. PS: staking v4 with full feature implementation is being used for Battle of Stakes Testing Campaign.
  [NOTE ADDED ON 08.04.2024]

Phase 2 → Epoch n+1: Redistribution of Shuffled-Out Nodes and Waiting List Adjustment

• Objective: To refine the waiting list management, ensuring fair opportunities for all validators.
• Implementation: Nodes shuffled out from the active validator set are moved to the auction list, not automatically to the waiting list.
• NOTE: As of Phase 2 of Staking V4, the network will have a total of 1280 nodes in the waiting validators list, distributed across 4 shards (320 nodes per shard - groups of 80 nodes per waiting epoch). This is a decrease from previous total of 1600 nodes (400 nodes per shard - groups of 80 nodes per waiting epoch). Previously, nodes would wait for 5 epochs before transitioning to the ‘eligible’ status. Now, while the transition to the eligible state still requires 4 epochs, the additional epoch has been shifted for nodes when in auction list.
  
  

Phase 3 → Epoch n+2 : Full Implementation of Soft Auction Mechanism and Node Limit Adjustment

• Objective: To fully implement the soft auction mechanism and adjust the maximum node count for each validator based on the top-up threshold.
• Implementation: Apply the soft auction selection process for all validators, adjusting the top-up threshold based on the average top-up in the network. This threshold determines the maximum number of nodes a validator can use to qualify for consensus, number which is being dynamically adapted.
  
  

What is the Soft Auction Mechanism?

The soft auction selection mechanism is designed to dynamically select nodes from the auction list to be distributed in the waiting list. This process is based on the top-up of each validator’s nodes, aiming to maximize the participation of different validators and their nodes in the network. Here’s a step-by-step breakdown and some examples:

Step 1: Assessing the Auction List

1. Auction List Compilation: Nodes that are not currently active (i.e., in the eligible list) are placed in the auction list. This list includes nodes from validators who have staked more than they currently need for their active nodes.
2. Validator’s Total Top-Up: Each validator’s total top-up is calculated. The top-up is the additional stake that an owner/delegator has put into their nodes above the minimum required stake.

Step 2: Calculating Top-Up Per Node

1. Minimum and Maximum Top-Up Per Node: For each validator, the mechanism calculates the minimum and maximum possible top-up per node. This calculation considers different scenarios where either all or only a portion of the validator’s auction nodes are selected.

• Example: Consider a validator with a total top-up of 3000 EGLD spread across 4 nodes (1 active and 3 in auction). If all three auction nodes are selected, the top-up per node would be 750 EGLD (3000 / 4). If only one auction node is selected, the top-up per node would be 1500 EGLD (3000 / 2).

Step 3: Determining the Qualified Nodes

1. Setting a Top-Up Threshold : The algorithm gradually increases a threshold for the minimum required top-up per node. This threshold is determined by gradually increasing the minimum top-up from the lowest possible value to the highest, in steps, ensuring the available slots are filled.

• Example: In this node selection algorithm, we incrementally raise a threshold within a specified range to filter nodes from the auction list. At each step, we compare the number of qualifying nodes against the available slots, and when we first encounter fewer nodes than slots, we revert to the previous step’s threshold. In cases where there are competing nodes with same top-ups a XOR-based sorting method is applied.

2. Qualifying Nodes for Selection: For each validator, the system calculates how many of their auction nodes can be qualified based on the current threshold of top-up per node.

• Example: Consider a scenario with a top-up (average) threshold of 1216 EGLD and two validators. The first validator has a single auction node with a top-up of 1222, exceeding the threshold and hence qualifying. The second participant owns three nodes in total—two in the auction and one active—with a collective top-up of 2555 EGLD. When considering all three nodes together, the average top-up per node is 851 EGLD, disqualifying them due to falling below the threshold. However, if we select only two of these nodes (one from the auction and the active one), the average top-up per node rises to 1277 EGLD, making both nodes eligible. In this example, first validator qualifies with one node and the second validator with two nodes.

Step 4: Final Selection

1. Sorting and Selecting Nodes: The qualified nodes are sorted based on their top-up per node, and the nodes are selected to fill the available slots. If multiple nodes have the same top-up, the selection is random but deterministic, based on a XOR operation between the validator’s public key and the current block’s randomness.

• Example: The final selection might include one node each from Validator1, Validator2, and Validator3, based on their qualified top-up per node and the available slots.

Conclusion

As we approached the Vega upgrade to Staking Phase 4.0, our primary objective is to enhance network decentralization and address the Nakamoto coefficient challenge. This upgrade is a strategic step towards the fostering of a more evenly distributed staking landscape. Phase 4.0 introduces mechanisms to incentivize smaller staking providers and diversify participation, that’s why it is important in our roadmap for a progressively decentralized architecture, setting the stage for future iterations that will further this goal.

Central Challenges and Proposed Measures:

1. Limiting Provider Dominance: With the MultiversX Network already hosting over 100 distinct staking providers, we are well on our way to achieving a decentralized ecosystem. Our goal is to further enhance this decentralization by encouraging the growth of professional staking services while preventing the dominance of a few entities. To approach this intention, we propose introducing protocol-level limitations on the number of nodes a staking provider can have. The first cap that we propose is of 50 nodes. How will this impact the actual staking providers? Staking providers that are above this number will simply not be able to add new nodes, but won’t get their surplus nodes removed by the protocol.
2. Supporting Small Providers: These limitations aim to reduce the growth of large providers and create space for smaller entities to flourish. It’s about creating a balance between competitive growth and maintaining a healthy, decentralized ecosystem.
3. Ensuring Network Security: From a security standpoint, these measures are designed to discourage any single provider from gaining excessive influence, which could lead to malicious activities. High top-up values for nodes act as an economic deterrent, aligning with game theory principles to prevent abuse.
4. A Call for Community Engagement: We recognize that these measures may seem somewhat restrictive, yet they are proposed in the spirit of preserving the network’s decentralized nature. This is an invitation for the community to engage here in a robust discussion. Your insights, perspectives, and suggestions are very important in shaping a more equitable and secure staking environment.

[davyegld]

The 100 staking providers were mentioned, but not the hundreds unique validators present.

Wouldn’t it help the decentralisation of the blockchain to force validators to have a minimum top-up stake to remain active?

Now there are many validators with exactly 2500 EGLD, what if the average top-up were counted and validators were forced to keep at least that threshold?

This way they would increase their EGLDs per node, or become staking providers to receive EGLDs from the community. Time could be given so that if the top-up drops below 10/20% average on the blockchain they have X amount of time to add EGLDs or become validators and receive them from the community (less top-up would give more APRs).

Example:

• Average top-up of 1000 EGLDs
• Validator with 2500 EGLD and 0 top-ups
• The system flags it as potentially removable
• Validator has 30 epochs time to get to the average top-up -20%, so 800 EGLD
• Or he becomes a staking provider and receives the 800 EGLDs from the community, which would stake for the higher APR

[stew]

Worst decision ever, this will lead a lot of independent validators to leave the ecosystem, this will lead to poor decentralization of the network. This decision is in favor of big players, not to decentralization. Unfortunately with this update perhaps i will leave the ecosystem as the greed of big staking companies has got you. Strat the auction with nodes from staking companies that have over 50 nodes. Or increase the stake gradually , predictable , not like this trough auction , you can increase like this: for year 2024 - 3000 egld /node, see how it goes, 2025 - 4500 Egld /node, 2026 - 5000 egld / node, and in 2026 make a DAO and see how you proceed from there.

[drmelone]

Okay I still have trouble understanding some parts of this, but I have grasped the general concept

The current proposal has one issue in my opinion.
The auction is a great self-governing model for consensus and chain security.
Especially in the far future, when more shards may be added and thus more nodes are required, the auction process, together with the wave of new nodes will naturally make the required EGLD stake become lower.

As more and more nodes enter the network because more shards are added, the average stake per node gets lower and thus the EGLD required to “win” the auction will become lower.

Awesome idea.
BUT the 2500 egld minimum staking requirement remains active with staking phase 4. Why? I do not like this. This kind of goes against the idea of the auction model, which is supposed to determine the required stake indepedently, automatically, by itself, no?

[drmelone]

Strong disagree. If the network wants to expand, it needs more nodes.

And you can’t increasingly ask for more nodes AND for more stake. EGLD is limited in amount. It will not work.

[live]

Can you explain a bit why do you think that will lead to poor decentralisation. How your suggestion of just increasing the cost of every node improves anything? The main goal of the auction is to make it permissionless, not increase the cost of a node. Right now someone has to leave for new validator to join, thats not open market. Yes, that will increase the cost of nodes, but thats the only way to make it open when you have limited spots for nodes.

[oldtimevalidator]

I propose that the auction for nodes start when there are more than 80-100 nodes in the queue (at atom you have to wait 2 years before you can catch a node), and the beginning of the auction to respect the principle of decentralization and increasing the nakamkto coefficient should be with batches of 10 nodes each, taken from providers with over 100 nodes (we are also talking about legacy here). The rationale is not to concentrate all the voting power (for dao) in the hands of a few providers. Series of 10 nodes will be auctioned until the list reaches 80 nodes again. The extra top up that will be placed in the case of the current auction system will only lead to a drastic decrease in APR, respectively the migration of stakerikor to other ecosystems that have attractive APR in the coming years (eg: aptos, sui). Let’s not forget that there are already providers with apr around 6%, which is totally unattractive. If we delude ourselves that users come for tech, it’s time to get out of the bubble a little, users come for apr and healthy growth. We have to take into account that the market is being educated and users are looking forward to new ecosystems (scroll, 5ire, aleph zero, base, etc.) that will take a good part of the users’ funds. I conclude by saying that we should not be concerned with the auction, but with increasing the number of transactions and conserving apr, maybe in the future by decreasing the number of active nodes if we do not have enough transactions

[robert]

@davyegld

Now there are many validators with exactly 2500 EGLD, what if the average top-up were counted and validators were forced to keep at least that threshold?

But this is exactly what Staking Phase 4.0 does. There is no more queue, so there is no infinite waiting time. Auction system works in the selection from “auction list” to “waiting list” and this selection is done via the top-up per node for each of the validators. The top-up counted is made in such a way to help the validators, so it will try to rebalance your topup automatically in a way that is best for the validators.

The system will show out that a validator node will not be selected as eligible node, it can stay on the network, and it has as much time as he wants to accumulate more eGLD in the topup and to enter as eligible. No need to get the node out and put back into it, only to put more eGLD in the system. Also, any unique validator can become a staking provider with a few commands, thus accepting delegations from users and increasing the topup.

So everything you explained and more is already implemented in the Staking phase 4.

[robert]

We waited more than 2 years with the auction model. It is quite the time to implement it.

And independent validator can become a staking provider and accept more funds to be delegated to him, thus increasing the topup and to be selected as eligible. Also the auction system runs at the end of every epoch, so you might be chosen even now, as in some days the minTopUp on the auction list can be smaller than on other days. After a few months everyone will tend toward the average topUp of the whole ecosystem.

So, just open up your node to accept delegations - become a staking provider and you will be able to gather enough delegation to be part of the ecosystem even after the activation of staking phase 4.

The decision is favoring small to medium players.

[drmelone]

You can’t have healthy growth and high apr in the same requirements

[robert]

@drmelone

ets lower and thus the EGLD required to “win” the auction will become lower.

Awesome idea.
BUT the 2500 egld minimum staking requirement remains active with staking phase 4. Why? I do not like this. This kind of goes against the idea of the auction model, which is supposed to determine the required stake indepedently, automatically, by itself, no?

Right now, after calculations we see that 1500 topUp per node will be required, that means roughly 4000 eGLD per node. So there is no need to scrap the minimum of 2500 eGLD per node as the demand shows that the market requirements is much higher.

As time goes one, we can lower the minimum 2500eGLD per node, depending on market conditions, demand for processing, demand for nodes and demand for shards.

The number of nodes will grow when there will be a growth in number of shards, which is the result in growth of demand. So when blocks are full on the 3 shards, a next shard will be added, making way for 500 more nodes.

[Adrian]

As long as the number of shards does not change, the number of nodes will also not change.
The idea for the adaptive system we are trying to create is that will adapt with the demand for nodes and demand for block space, where these influence each other. There is no need to increase the number of shards if there is no demand for block space as it will not be economically attractive for new nodes (which share the fees). So in order to keep the network performant and the incentives high enough that they are attractive we need some corelation between the two.

While having the limited number of nodes (for the level of blockspace demand we have right now) will not cause the stake per node to go below 2500 even if we enable it right now.

The change in the minimum threshold can thus be delayed.

[Adrian]

all nodes shuffled out will go through the same auction.
If any node no matter that it belongs to a staking provider (contract) or individual validator (wallet) does not meet the topup requirements decided through the auction, then it will not be selected.

[robert]

I propose that the auction for nodes start when there are more than 80-100 nodes in the queue (at atom you have to wait 2 years before you can catch a node

The solution @oldtimevalidator proposed does not enable an open free market which we aim for. It is still limited. Having 80-100 nodes in the queue is not a hard thing to have now, but an OPEN-MARKET is much more important.

APR increase will come as the demand for blockspace is getting higher. The number of users is increasing right now, more and more DeFi projects are on-chain and a lot of other products are definitely growing.

Building and supporting builders is a top priority. But staking phase 4 creates the primitives for a new OPEN MARKET for the validators and staking providers which will increase the demand of eGLD, as new providers will enter the ecosystem, with new users.

Technology is a key element to create innovative products which reach billions of people. That is a key driving force for people to build, and users will come to the ecosystem as we have more and more products built for them.

[drmelone]

@oldtimevalidator

I conclude by saying that we should not be concerned with the auction, but with increasing the number of transactions and conserving apr, maybe in the future by decreasing the number of active nodes if we do not have enough transactions

Okay so you agree with what @robert said?
You are not concerned with the auction or staking phase 4 for that matter.

Your main concern is preserving APR by bringing more usage to the chain.

@robert

APR increase will come as the demand for blockspace is getting higher. The number of users is increasing right now, more and more DeFi projects are on-chain and a lot of other products are definitely growing.

Building and supporting builders is a top priority.

I think future discussions about a transaction fee market or generally bringing more adoption for that matter are more goal oriented - staking phase 4 has little to do with this then, you say it yourself even

[drmelone]

As far as I understood, validator = staking provider in terms of what they need to have to be eligible for consensus (topup stake).

So for validators that don’t allow delegation, they will need to put more money into the game.

[davyegld]

Ok thanks this was the missing puzzle piece I didn’t understand.

So the auction will be for all nodes waiting and we’ll not have a differentiation between nodes in queue and waiting.

So for example we’ll have 400 nodes validating per shard, and others that can enter. So we’ll remove the 3200 nodes limit?

[drmelone]

3200 nodes limit will not be removed afaik

@robert you said “a new shard will enable 500 new node spots” in your response to me earlier.
why 500?

Isn’t it right now: 400 active nodes, 400 in “waiting list” - so 800 per shard basically?

[steak4all]

I have large concerns about phasing 4. I hope I am just not understanding it correctly. As I see it now, this will strongly lead to MORE centralization and heavily favor the top 20 staking providers while pushing all small SP and independents out of business. In addition it will be virtually impossible
for any new staking providers to enter the market.

Let’s look at these concerns in brief.

The vast majority of stakers, I’d guess over 95%, only use one metric to decide where to stake their EGLD, and that is APR. No amount of special staking rewards, nfts, promotions, give-aways, marketing or anything else will change this fact. I know that’s not the ideal, but that’s the reality, so lets focus on reality for the sake of this post.

The top 20 staking agencies today, Dec 12th, control 1,938 of the 3200 total nodes. That’s 60% of all nodes.

If you are in the top 20, or even have over 8 or 10 nodes, you can adjust your APR and top-up in many ways that ALL the node operators with less nodes simply can not do. This means that the top 20 staking providers will be able to adjust their top-up and node count to easily squeeze out the smaller operators, and the smaller operators will have no recourse, and forced out of business.

So this follows, early in phase 4 (the first days) a line will be made: those large providers that are secure in the system, and all other providers will be unable able to compete and forced to close, and there will be next to nothing they can do to prevent that.

Sorry for caps, but I want to make this clear: STAKING PHASE 4 WILL ASSURE CENTRALIZATION AND REDUCE DECENTRALIZATION.

Once that line in the sand is drawn, ALL of the staking revenue for ALL of the providers (most all of them only different because they are small not large) will be cut OFF. This means that the APR for all providers unable to meet the average top up will be cut off to 0%. This follows that all of their
stakers will leave their staking providers. How would a staking provider with 0% APR EVER be able to reach the necessary EGLD deposits to become an active rewarded node?

Staking phase 4 as I see it, (and I honestly hope I’m just seeing it wrong), will quickly lead to there only being 20 or so SPs and no new competition will have any chance at all in joining unless they are backed by very large investments, of at least over $100,000 US, if not more.

Already in staking phase 3.5 today, the larger provider you are, the better off you are. Let’s look at an average 10 node SP. Due to their active node count, they have a much larger share of the rewards coming in. They could set a fee of 12% say, and their APR would be 7% for their stakers. Compare
that to a small node operator like steak4all, with two nodes, even if we were to set the fee to 0%, losing money paying the cloud fees to keep in operation, we would not be able to return the same APR. In phase 4, its even more dire, because now, if the don’t meet the top 10% of delegator numbers according to top-up, then ALL rewards will stop, which directly will quickly lead to exodus of users, and there is 0% chance they’ll ever be able to then attract enough users to become rewarding again.

Staking phase 4 seems like it’ll wipe out most – the majority – of all staking providers. Only the giants will be able to adjust their top up by scaling nodes accordingly to remain in rewards, permanently freezing out the MAJORITY of staking providers from getting rewards.

This appears to be a disaster in the making, and I don’t know why more people don’t see this.

I’ve been looking forward to staking phase 4 as the first chance I’ve had to grow my staking provider since we launched on Day One in 2021. In the last two years, any time I tried to bring a 3rd node online it would crater our APR so many stakers would leave, causing us to fall below the required EGLD
amounts to sustain the expansion, so we had to bring the new node offline. But now I’m dreading Phase 4, sadly. I know that if I don’t make the top-up cut-off in the first week of its implementation, then the chances of ever being able to reward all of our delegators will be next to zero, and it follows that’ll we’ll likely be pushed out of business.

If anyone has any questions or would like to point out errors in the post, please do so, because I’m really hoping I’m wrong.

Phase 4 will wipe out most small providers and will further entrench the Top 20 staking providers (that own over 60% of the network currently) to basically become gatekeepers closing the gate on all smaller providers and ALL new providers from even entering the ecosystem. I don’t see how this won’t be a disaster.

How will the top largest providers maintain the stranglehold on rewards?
Due to their size, large node operators can easily scale their node number to maintain the highest top-ups, perpetually assuring their rewards, will forever freezing out the MAJORITY
of node operators from making ANY rewards. With no rewards, users will quickly leave the providers not making the needed top-up levels, and with 0% rewards, they will not be able to increase the amount of EGLD delegators of their service, directly leading to them going out of business.

Phase 4 will simply change the staking to a monopoly market whereby only the largest existing providers can compete, and all others will not be able to compete and be forced out. How does that improve decentalization?

How will staking providers with 0% APR rewards keep users?
They simply will not be able to, and they will quickly lose all their users, and never be able to meet the levels of top-up required to stay in operation.
No amount of NFTs or community spirit or anything will change that fact. The vast majority of users, I’d guess over 90% at least, only judge a SP by their APR rewards, and have no idea about their X presence or token or NFT offerings etc etc. Phase 4 won’t drive further engagement, it’ll drive small providers out of business before that happens.

What effect will only the monopoly staking providers earning rewards have on the staking ecosystem?
It will mostly kill it. And with that, kill innovation, competition and extra offerings/rewards developed by staking providers.

How will a new staking provider launch in the Phase 4 system?
It will be next to impossible and require massive monetary investments only available to larger corporations to effectively launch a new staking provider under phase 4.
Why? Because rewards will only go to the largest node operators, and until a new staking provider is a large node operator, they’ll be effectively shut out. It will required what (the exact numbers are impossible to say at this time) at least 7000 EGLD to start a staking agency that is
rewarded for staking? Today that’s $435,000 US. Compare that to 1250 EGLD that it is now. Which is better for fostering growth? The answer is clear.

This again will be very bad for decentralization and will have the opposite effect: creating a monopoly market where only the established big boys will survive, which massively will hurt the growth of EGLD and user engagement.

But you say: Steak4All has relatively high top up and will be able to be rewarded, so why is this bad?
After phase 4 is implemented, the top-up levels needed for us to survive will change the same day. Just considering the top 20 staking providers, controlling 60% of the nodes today, will be able to adjust their node-to-topup ratio easily, reaching the reward level without issue. This means that the majority of all
rewards will go to the largest providers, and smaller providers will be frozen out of rewards.

Think about phase 4 another very basic way: 1600 nodes of 3200 are rewarded. 60% of the 3200 nodes are controlled by the top 20 largest providers. So, 1920 nodes are controlled by the Top 20 largest providers. The largest SPs have enough delegators to ASSURE they’ll be in rewards adjusting their node count. Thus ALL of the rewarded nodes, 100% of the rewarded 1600 nodes, will potentially be controlled by the top 20 SPs, and 0% rewards will go to 80% of SP’s operating! Please tell me I’m wrong.

Staking Phase 4 appears to achieve directly opposite of the stated goals:

1. Limiting Provider Dominance: The opposite will happen, as discussed. The ‘possible’ node limit of 50 nodes doesn’t even matter much because it would not effect existing providers, and as mentioned, most of top 20 operators have over 50 nodes currently, so they already have enough staked EGLD to have monopoly control over staking rewards. Staking phase 4 as proposed will effectively ASSURE the domination of the largest providers.

2. Supporting Small Providers: It is plain that a new provider will effectively 0 chance of being able to enter the monopoly-controlled SP market, no matter what additional services, offerings, bonuses, NFT or token rewards they offer. They’ll never be able to reach the minimum top-up level for rewards
   while receiving 0% rewards (what delegator would sign up for 0% rewards for an indefinite amount of time, probably forever, when the monopoly providers are offering 7% APR the same day? That’s unfair for delegators. All delegators should earn interest, not just the delegators to the largest top SP’s). Phase 4 will effectively close off EGLD staking for ALL new providers and most small
   providers.

3. Ensuring Network Security: As shown, staking phase 4 will lead to massive centralization to a few mega-providers, putting most small providers out of business, thus hurting network security, directly opposite of the intended goal.

The purpose of Agora is to have a dialogue, so we hope one is opened here. Most small providers will not be able to survive Phase 4, as proposed, if what I’m saying here is correct.

Changes could be made to this plan. Adjustments could save the entire design. If we can agree that as proposed, staking phase 4 will be a disaster, then I’d be happy to work on offering solutions.

If you disagree with the premises forwarded in this post, please bring them up, or ask questions so we can further figure out if they are correct or not. As said at the beginning: I hope this post is entirely wrong, because as it appears to me, Phase 4 will accomplish the opposite it sets out, a massively further centralization – not decentralization, which was what we’ve been hoping for for over 2 years now.

Thank you for reading!

[drmelone]

@steak4all

Limiting Provider Dominance: The opposite will happen, as discussed. The ‘possible’ node limit of 50 nodes doesn’t even matter much because it would not effect existing providers, and as mentioned, most of top 20 operators have over 50 nodes currently, so they already have enough staked EGLD to have monopoly control over staking rewards. Staking phase 4 as proposed will effectively ASSURE the domination of the largest providers.

I see 9 providers (maybe I even missed some) that have (way) more than 50 nodes right now.

They will need to close them.
They will have way too much top up.
APR will be super low.
Users will leave
Where do they go?
Elsewhere, where APR is better.

@steak4all

what delegator would sign up for 0% rewards for an indefinite amount of time, probably forever, when the monopoly providers are offering 7% APR the same day?

Okay but that’s not different to our current system. With the current system you’ll be in waiting list for a year or whatnot and with the new system you can start being a part of consensus as soon as in a few days.

@steak4all

As shown, staking phase 4 will lead to massive centralization to a few mega-providers, putting most small providers out of business, thus hurting network security, directly opposite of the intended goal.

I understand all of your critique, but here- how? Even if all providers have 50 nodes, that would be 64 providers. And we have MANY small providers and very little providers in the 50 node ball park.

The only thing I agree with you on is that the new system doesn’t make it easier for small providers. But it also does not make it tougher in my opinion. It’s the same situation in the end.

(PS: Your post was talking about the same three points over and over again, I think you addressed each point 4 times each)