Polkadot Staking Changes: Progress & Timeline

Governance
22

I am myself a nominator and will get affected by the lower numbers proposed here, but I see some benefits to it. I will share a few thoughts here from my personal standpoint:

Ultimately staking reward is not really a reward or value being generated but a tax imposed on those who do not stake, to pay those who do stake and put their funds at the risk of being slashed. If everyone who holds DOT would be staking, no one would be making any money and only a mild continuous denomination change would happen. So perhaps this can help remove the idea of “reward” and replace it with “tax on liquid DOT holders”.

The first thing then to understand is that this tax has been quite high since the genesis of Polkadot, and moving too much capital around. Moreover, given that it has been pro-rata to how much stake you have, it has likely made the “(already) rich only richer”.

I understand that that as stakers, this all comes off as “we receive less rewards”, but bear in mind that this also means that all those stakers who have larger bags than you and me also receive less rewards, and their cut will be even larger. And previously, even if you were staking, you were also paying this tax to provide the income for stakers larger than you.

Finally, this amount of tax that was being moved in the ecosystem, which we may call “the cost of Polkadot’s security” was nowhere near close to the revenue and inflow of DOT. At previous prices, Polkadot would have been taxing up to $500m per year to compensate its stakers.

So the previous order of things, before the hard cap, could be summarized as:

  • Polkadot is over-spending on security, which leads to:
    • at best, more and more centralization of DOT holders (making the rich richer)
    • at worse: liquidation of DOT, as those who accumulate excessive amounts DOT are more flexible on leaving their position

While with the hard cap, and the rest of the steps proposed in the DAP proposal, at the very minimum, we are slowing down all of these negative trends, and at best we will have a chance at reversing some of them. Of course, this will come at the cost of less DOT distributed to stakers, but I personally hope that the protocol will have a higher likelihood of better DOT price action with these changes in place.

All of this is about reducing supply and not increasing demand, and indeed increasing demand is also necessary. I hope that Parity will share exciting news to the community about that soon and in the Web3 Summit.

Specific note on why nominators are seemingly taking a larger share of the “reduced spending” (even though so far I have argued that it is not a purely bad thing): We can think of it this way – the only way for Polkadot, as a protocol, to reduce its security spending is to hold less DOT at risk of slashing, and in return pay less for it. Between nominators and validators, it would only make more sense for validators to hold the majority (or ideally all) of the slashable DOT, ergo the introduction of the 10k min self-stake. So the only portion that we can remove from the slashable pile is the nominator stake, ergo nominator stake becomes un-slashable. This will also make nominator stake unbond-able in just 2 days vs. 28 days for validator-stake, which is a fair exchange.

23

I propose a radical simplification of Polkadot’s reward and treasury model:

  • Validators: 12%

  • Nominators: 1%

  • Treasury: fixed at 1 million

  • All remaining issuance: burned

Rationale

The current system is fundamentally inefficient in capital allocation.

Treasury spending has demonstrated:

  • lack of rigorous evaluation

  • weak accountability

  • poor capital efficiency

A significant portion of funds is either underutilized, misallocated, or directed toward low-impact proposals.

At the same time, governance is increasingly dominated by large voting blocs, which further distorts incentive alignment and reduces the effectiveness of treasury distribution.

Why This Change

This proposal enforces discipline at the protocol level:

  • Drastically limits treasury misuse by capping available funds

  • Aligns incentives toward long-term value through deflation (burning)

  • Prevents governance capture from translating into unlimited spending power

  • Forces higher proposal quality, as funding becomes scarce and competitive

Conclusion

If the system cannot reliably allocate capital efficiently, then it should allocate less.

Burning excess issuance is not wasteful — it is a mechanism to protect value in the absence of effective governance.

24

Okay, even agreeing with you and understanding your point of view, doesn’t a 70% return for the validator seem excessive to you? They’re going to put up 30,000 DOT and receive 21,000 DOT per year just because they’re the only ones taking on the risk of slashing? You could argue that they also face the risk of having to lock up 30,000 DOT, but in the end that’s a risk that regular investors and nominators also share.

25

“And 18 million DOT for the treasury? If they’ve been responsible for the drop in DOT’s price due to unnecessary spending and waste on projects and ideas that didn’t work out. In other words, we want to reduce spending on nominators, but at the same time we’re giving 18 million to the treasury to spend and a 70% APR to validators. What I’m seeing here is all the money going to Parity, its subcontractors, the people who work for them, and the outsourced teams. And the validators are very likely people from the Parity team. Honestly, I see it as draining all the money to the team—because the only expense I’m actually seeing being cut is that of regular people staking their DOT.

26

It is, but:

  1. I think 70% is viable if you are the only validator who has say 30k and everyone else has 10k. Maybe I’ll build a playground around @jonas’s formula (linked above) for self-stake to explore this more. While it is a generous one, it is not that good where it is spoiling validators with crazy APY.
  • I actually don’t know where you get this 70% number from. It is not mentioned anywhere here.
  1. This APY for validators peaks (in terms of pure return) at something like 30k DOT, grows in a diminishing way up to 100k DOT, and after that you get no extra return. The average validator today likely has millions of DOT in self stake (behind anon nominators), and the 8% they earn on that is much greater cost to the protocol than the 20/30/40% they might make on 30k DOT self-stake. So the face value of the APY might look very high, but note it is bounded.
  2. These parameters can definitely change once we can pay them a fixed stable income, and at that point their DOT payouts can be made both less + vested, since they can pay their costs from the stable part.

This is completely a misunderstanding. The reserve part of the DAP is basically un-touch-able by the treasury, and will not be spend in such ways. This DOT will be saved and the only reason to spend it would be to pay for future staking rewards in any form (DOT or stable).

And as you can see, with the halving of the issuance every two year, the protocol has a good chance of having a multi-decade runway: Save DOT today in the DAP while the issuance is high, have higher guarantee to be able to pay stakers a reasonable amount in the long run.

27

The return probably has to be proportionate to potential price swings. 30K DOT last year was ~110100 USD vs ~35000 today.

28

You’re focusing on the wrong thing.

  1. Whether it’s 70% or not doesn’t matter. The real issue is that the model still incentivizes splitting stake across multiple validators. Returns are “bounded” per validator, not per entity — so any rational actor will just run more validators.

  2. That’s basic game theory. So instead of improving decentralization, you’re just making it look decentralized while concentrating control.

  3. The same goes for the DAP reserve. Calling it “untouchable” doesn’t remove risk — it just pushes the problem into the future without any clear framework for how that capital will actually be managed.

  4. Right now, there’s still no entity-level constraints, no decentralization metric, and no real mechanism to prevent Sybil behavior.

So this isn’t really a misunderstanding — it’s just an incomplete design being treated as if it’s solved.

29

This is why I’d LOVE this PoP consensus to make its way to fruition. You’d know who has what and if single validation pools are actually different entities or not?

But I’m not expecting that to come any time soon. All of the decisions for OpenGov are ultimately decided by 10 people who have a tremendous amount of DOT in holding through their delegations.

There’s maybe 300 separate entities that vote on Subsquare/OpenGov regularly and none of the other 290 votes even matter. I wouldn’t even be surprised if they are all ran from one account? How would you even know? They always vote lock-step with each other shutting everything down, might as well be the same with every validator on Polkadot/Kusama.

30

This is a very important topic.

Even in the current situation, the actual level of decentralization is much lower than what is being claimed.
For example, this screenshot shows only some of the operators that control multiple validators.

image
image937×301 32.5 KB

Also, the address 13FzGLWoueKvUqFePiJgvFYWhH5KckHGtVBXvAX7SBtVZbXu recently created several dozen new validators, many of which are now also in the active set, without any clear identifiers.
For example, do you know who is behind this address and these validators?

If nothing changes, we will end up with only the illusion of decentralization, while in reality the blockchain will be controlled by just a few teams or companies.

31

Thanks for the write up. Which stablecoin(s) and are nominators still a factor in determining the active set in this future model?

Thank you

32

The decentralization concern raised by @LEGEND, @dandan, and @StakeUp is the most important unresolved issue in this thread.

I agree you can’t cryptographically prove two validators aren’t the same entity. But you can make it much easier to identify genuinely independent operators using information that already exists. A few practical signals:

Infrastructure distinctiveness

Independent operators who own their hardware can demonstrate things that multi-validator clusters on shared hosting cannot: unique ASN not shared with dozens of other validators, hardware-specific benchmarks that differ between nodes (identical profiles across “different” validators suggest identical infrastructure), and distinct system configurations that reflect individual operator decisions rather than templated deployments.

On-chain behavioral correlation

Multi-validator operators tend to show correlated behavior a scoring system could flag: simultaneous session key rotations, identical commission changes at the same block, correlated downtime events, identical payout timing. This data is already on-chain. A tool scoring validators on “behavioral independence” would give nominators a measurable decentralization signal.

Self-stake incentive and Sybil economics

@dandan correctly identified that the current structure rewards splitting stake across validators. If the marginal self-stake incentive diminished beyond a detectable entity-level threshold, the economic reward for Sybil behavior would shrink — not eliminate it, but reduce it.

Identity depth over identity existence

Not just “set an on-chain identity” but verifiable depth — infrastructure documentation, public monitoring (e.g., Crunch bot payout rooms), track record across networks, governance participation history. An identity that exists across multiple independent platforms over time is harder to fabricate at scale than a display name on-chain.

Why this matters now

If the active set shrinks to 250–300 as proposed, every slot becomes more valuable. Without a way to prioritize genuinely independent operators, the reduced set will consolidate toward well-capitalized entities running clusters — the opposite of the stated goal. A community dashboard scoring validators on independence metrics wouldn’t require protocol changes to start, and would create organic market pressure toward real decentralization.

Full disclosure — I operate SafeStakeFirst, a single bare-metal validator on private infrastructure with 11,850+ DOT self-stake. I have direct interest in seeing independent operators valued. But the point stands: if the network can’t distinguish between 300 independent operators and 30 entities running 10 validators each, the active set size is theater.

33

I am addressing a few things that have been raised by a few people throughout this post.

The proposed self-stake incentive system does have marginal returns to scale to prevent a few validators with very high self-stake from soaking up all the rewards. The mechanism is designed to target a uniform distribution of stake across all validators at an optimum of T=30,000 DOT. Importantly, it does not disincentivize stacking up more self-stake, nor does it push validators that currently hold more DOT to get rid of it.

This does imply that it is profit to run multiple nodes. But this has always been the case: every operator already has an incentive to fill all active slots to maximize returns. With this system, we ensure that it is at least quite capital intensive, giving a decent economic security per validator.

Of course I agree that decentralization is very important, but it cannot be solved purely with mechanisms hardcoded at the protocol level. The core problem is that we can never reliably desybil operators. Even if we could, whoever passes such a process can still be entirely different from the person actually operating the nodes. So, trying to approximate decentralization with on-chain metrics just incentivizes operators to game those metrics more sophisticatedly, which ultimately hurts transparency. We saw similar issues with the 1k-Program: For example, giving higher scores for unique node locations led to operators spoofing their locations. It took a dedicated team to constantly monitor and investigate telemetry data to fight this. This cannot be automated to a degree that it is helpful.

On top of that, even defining a “decentralized validator set” is nearly impossible. You get a different answer from everyone you ask. How many nodes may an operator run and still contribute positively to decentralization? Are 2 nodes on the continent of Africa by a single operator “more decentralized” than 2 nodes by distinct operators in Europe?

This is why nominators are so important and a key part of the system. It is ultimately nominators who must do the work of aggregating information about operator identity, geography, and conduct to arrive at what most would consider a decentralized set. Nominators should ideally engage personally with operators, assess whether they are small independent actors running nodes responsibly, and let that inform their nominations. No protocol mechanism replaces that human judgment layer, it can only support it. We’ll be sure to continue evaluating additional tools to support this process (such as PoP), but the main point stands that human judgement needs to drive the selection.

34

But do nominators have sufficient incentives to promote validator decentralization through their nomination choices?

From my observations, most nominators are primarily focused on maximizing their staking returns. Selecting a sufficiently decentralized set of validators does not appear to be a primary consideration. In addition, some large nominators, such as centralized exchanges like Binance, tend to nominate validators they operate themselves that charge 100% commission.

Will these issues be addressed in any upcoming changes?

35

In my view, the most effective way to discourage operators from running multiple validators and multiple identities is to fundamentally rethink the validator reward and election system

If our real goal is to have approximately 300 validators operated by 300 different entities or individuals (or any other number), rather than a handful of large players controlling hundreds of nodes - then we may need to move away from the current Phragmén model. A better approach could be to let nominators select just one validator per nomination (with the same amount of stake), and distribute rewards proportionally to the stake each validator receives, similar to how it works in many other blockchains

This change would remove the economic incentive to run multiple nodes. Validators would be motivated to focus all their energy and resources on operating a single, high-quality node instead

To ensure smaller validators can still sustain themselves, we could introduce a modest base reward, for example, a fixed amount per month with the remaining rewards distributed proportionally according to stake. This would prevent validators with very low stake from earning almost nothing, while keeping the system fair

It’s widely acknowledged that the current validator set has some significant issues. The existing model already encourages operators to run multiple validators, which has unfortunately led to many using different identities as well. With the upcoming changes, a single large entity could even more easily run dozens of validators by simply providing the required 10k self-bond for each

For these reasons, I believe a more fundamental reform would be healthier for the long term. At the moment, it feels like we’re reinforcing a system that is already showing clear weaknesses