Figment

Resources

Why We Voted ‘YES’ on Polkadot Governance Referendum # 31

disclaimer: This article is mainly in reference to Polkadot Referendum 31 and the information herein is current until new governance changes take effect to change the parameters that the referendum impacts. Thanks to my team members at Figment for reviewing the article and to Will Pankiewicz at Parity.io for editing this article to clarity.

In a nutshell:

Polkadot’s Referendum 31 raises individual minimum DOT staking requirements from 40 to 80 DOT per nominator due to acute network architecture limitations.

We support this position because it will ensure network stability for Polkadot over the time necessary to upgrade t network infrastructure in accommodation of filled nominator maximums.

In this article, we’ll go into our support for the referendum, after evaluating motivations for its proposal, and why it’s important that all stakers attempt to participate in governance proceedings for the platforms on which they stake.

How does this affect those who want to stake with us?

  • If you are choosing to nominate any one of our addresses, you will now have to nominate a minimum of 80 DOT to our nodes.
  • If you are currently nominating less than 80 DOT to our validators, your nomination will eventually be chilled out of the set – a process that will take approximately 28 days.
  • If you are nominating much larger quantities of DOT, please feel free to get in touch with us so that we can discuss the benefits of staking with Figment Prime, but this referendum will not affect your nomination.

Background

As an infrastructure provider, we play an active role determining the future of protocols. Through governance processes, we can help influence the future of protocols. We review governance actions closely as protocols work to improve over time through proposed changes.  However, from time to time, a governance proposal can upset the wider community of a protocol or network.

Polkadot’s Referendum #31 is a perfect example of a governance proposal that upset the community when proposed. This article dives into why we supported a proposal that may appear to go against the interests of the wider community but we think will instead support the future of the protocol to the benefit of the community.

In line with Vitalik Buterin’s most recent blog post on Moving Beyond Token Governance, we think governance is a crucial element of Web 3, and we want to participate in it at all levels and all forms to ensure that we can provide well-decentralized improvements to the platforms we support – so describing our deductive approach here should be helpful to all those who choose to stake with a platform.

Details of Referendum 31

  • Referendum 31 raises staking nomination minimum from 40 DOT to 80 DOT
  • Referendum 31 was passed, and adopted, and nominators below the 80 DOT threshold will steadily be ‘chilled’ until they are no longer in nomination, opening up slots for validators above the 80 DOT threshold. 

The motion was proposed on August 5th, 2021 on Element in this message. It was put to Motion 103 at block height 6322630, accepted for motion at block 6332970, and put to Referendum 31 at block 6348546.

The proposal cleared public vote on August 17th and was enacted at block 6421146 – quantitatively a fairly quick enactment of 7 off-chain days in order to mitigate any further potential of instability in the platform’s current state.

Why was this Referendum proposed?

Polkadot’s consensus method was artificially capped at 22,500 nominators for the acting potential of the network, an amount that was reached through Polkadot Referendum 28, which also raised nomination minimums from 20 DOT to 40 DOT, which has fueled people’s criticism that minimums are being raised to push people out of the network.

Within the current cap, the nominator maximum was reached faster than was expected, and the system is at risk of instability. The current referendum, Polkadot Referendum #31, focuses on mitigating the instabilities to buy more time for the network and ensure stability before hitting maximum nomination.

A quick overview of staking on Polkadot

If you’d like a more in depth description, Gavin wrote a great post summarizing the nomination process on Polkadot, which has largely been left unchanged.

  • Stakers nominate up to 16 validators on Polkadot and 24 validators on Kusama to whom they would like to delegate their stake.
  • The stake is then delegated among these validators through the Phragmén algorithm to add to a validators election total, which in turn allocates stake toward the validator with the highest chance of meeting the minimum election necessary to participate in validation. This algorithm will usually end up only backing one of the validators chosen.
  • The next set of active validators are selected by the active set of validators at the end of the previous era using another off-chain Phragmén allocation. All elected validators have the potential to take part in the active set.
  • Click here if you want to learn more about how Phragmén works.

How do rewards work on Polkadot?

  • Each era (24 hours on Polkadot, 6hrs on Kusama) all validators from the active set are disbursed roughly equal rewards, regardless of total stake amount on an individual validator.
  • Validators then deduct their commission and portion out the rewards to those who nominated them.
  • Nominators then receive rewards proportional to the percentage of stake that their nomination comprises of the cumulative stake on a particular active validator during each era.
  • These rewards must be unlocked by any party from a particular validator pool. (We take care of this for you if you stake with Figment)

Why did we support the referendum?

Being that the situation around Polkadot’s present WASM architecture is a hard limitation on what the network can provide in terms of security and stability for the consensus mechanism – the decision is unfortunately a pragmatic one.

Yes, this leaves out stakers at the bottom – but due to current limitations, the choice has to be made between ensuring that the platform can continue to perform properly, or to ensure that the platform remains open to people at lower minimums. This proposal unfortunately encapsulates the reality of how protocol changes can affect all members of a community and their connection to the network.

Since we see a future for this platform, we want to see it grow. The only way it can do so if we take measures that are in the best interest of the network architecture first, then collective rewards second – leading to a conclusion in support of this proposal and knowing full well that in the long run, we want to see these minimums lowered.

As such, we committed a measurable amount of our tokens to the vote with a 2x conviction, meaning that our on-chain holdings will be tied up for 56 days, doubling the amount that was submitted. You can view our vote here and here – and a more in depth description of our analysis for taking part in the decision below.

Why enact Referendum #31?

Most importantly: This measure is intended as a stopgap until other network solutions are sought and implemented to accommodate both more nominators and validators within network architecture, at which point the staking requirement is to be lowered. Nominator slots were filled faster than anticipated, which could have led to memory allocation issues and resulting block production instability.

The network has a nominator maximum of 22,500 across 1200 validators, making for 256 active nomination slots per validator. The network has stayed at total maximum nominator count for the last two weeks – no more nominators have been able to consistently join the set.

How this affects the platform has to do with both the WebAssembly architecture of Polkadot and limitations of the safe memory allocation in off-chain computing of rewards and transaction allocation. Having all nomination slots filled puts stability of the network at risk and could result in both oddities in rewards distribution and instability in consensus. As a result, an urgent change needed to be made in order to ensure stability as the system continues to scale, leading to this proposal.

The recommendation in this scenario could be a number of options. The platform could artificially reduce the number of nominator slots available per validator- but this could lead to slot sniping and outsized rewards for those who are able to secure slots.

Leaving these slots open and raising the minimum required to take part in nomination allows for eventual at-volume use of the platform with slower uptake in nominators onboarded to staking. The latter option provides the platform with the time (approximately one era of 28 days) to expand capabilities for managing filled nominator slots.

In the long term, nomination minimums will ideally be lowered in order to accommodate stakers at lower points of entry for rewards – and if not the case, other solutions may be sought in the form of staking pools while slot availability scales. This will be done through the same governance mechanisms that have proposed the current referendum.

What is causing these issues?

One main reason has to do with nominators not completely understanding how the staking mechanism works within Polkadot. There are enough nominators staking who don’t understand what the Phragmén is, or how it disburses rewards to those who nominate. As such, some nominations have resulted in a ‘blind delegation’ that fills slots with inactive nominations.

In some cases, nominators aren’t earning rewards due to delegating their stake to oversubscribed validators at the present nomination minimum. In most cases, those who hold below the new nomination minimum of 80 DOT are not being added to the active staking set for each validator, leading to nominators that aren’t earning rewards. Additionally, Polkadot simply hit benchmarks faster than was expected, and the platform needs more time to scale properly and safely.

How does the referendum affect the protocol?

The intent of increasing the nomination limit is to decrease the number of nominators who access nomination slots, thereby decreasing likelihood of WebAssembly instability on the network. With a raised minimum, less people will meet staking requirements on the network with their tokens, and it will ideally increase overall network resiliency and security as people are more knowledgeable in nominating their stake to lesser staked validators.

As mentioned above, this is not intended to be a long-term solution – It’s clear that the popularity of staking on the Polkadot platform has introduced a free-riding problem where the system cannot accommodate all nomination slots being filled, particularly by those who are oversubscribed. Current network architecture cannot account for this. Eventually, the number of nomination slots will be drastically increased on the order of many times the current amount due to improvements in network architecture.

How does the referendum affect the community?

As mentioned above, this proposal means that nominators below 80 DOT will gradually be ‘chilled’ out of the active set of nominators to open up slots for incoming 80+ DOT nominations.  If you are one of these chilled nominators, this enacted proposal can feel a bit like an attack – it feels like consolidation of the network into only those who can take part.

However, the technical bottlenecks of the Polkadot platform and it’s WASM architecture are at fault here – indeed this is not the first time minimums have been raised to deal with potential stability issues, with the last time reaching as high as 200 DOT.

The net effect is that this proposal will lower the amount of individuals who can take part in direct staking on Polkadot and thus take part in on-chain consensus.

How might criticism be responded to?

Referendum 31 may seem to ensure only those who already hold large amounts of DOT can who get to take part in receiving staking rewards – and to a degree, this is true. Largely this proposal does nothing to affect large token holders – but it doesn’t necessarily change the dynamic that existed prior to the minimum.

Truthfully, the only faith that the community can have is in the legitimacy of the project and the roadmap for continued expansion that will eventually accommodate a larger number of token holders with lower minimums. If the minimums are raised again, it must be understood that this is an indication not of centralization but rather of platform popularity that will eventually lead to the expansion of the platform in accommodation of lower and lower nomination minimums, and a higher number of nomination slots.

Large token holders now receive a marginally higher portion of returns as split across the proportion of total stake, but since the original percentage that <=80 DOT holders were receiving is still quite low, the net effect shouldn’t necessarily be seen as a power grab – these larger token holders would be a part of the network no matter what, since they will consistently be above the minimum required stake.

Smaller token holders can absolutely still take part in governance proceedings using bonded DOT using any level of conviction to multiply the importance of their vote. Parity have mentioned that staking pools may eventually be launched to provide means to stake when below the individual nomination minimum.   

Additionally, if nominators would still like to participate in on-chain staking and remain within the Polkadot ecosystem, they can do so by staking on the canary network, Kusama, which presently has much lower equivalent nomination requirements.

Ethical position on the maintenance of our public-facing knowledge

We intend to have those of you who stake with us understand the full breadth of our decision-making process for the governance decisions to which we are party, so that you as a staker can proceed with the necessary information to take part in the same governance decisions that improve the networks on which you stake. 

As such, I wanted to summarize the evaluative process that we employ to examine the implications of governance decisions on any particular protocol. It’s important for community members to understand that decisions such as these are not taken lightly by members of the community, and that judicious attention is paid to what is best for the growth of platforms, even when a decision can seem, on it’s face, to be centralizing or malicious.

And who knows, if you learned something from this, maybe you should stake with us, 😉

Further reading and information:

  • A Polkadot Post-Mortem – A summary of instability issues Polkadot encountered at block height 5202216, where memory allocation issues resulted in overflow to the WASM and halting of the chain, which is the cause for caution resulting in both Referendum #31 and #28
  • Polkadot Governance – A summary of governance ethics on Polkadot.
  • You can follow me on Twitter here.


Stay in Touch