Validator Decentralization: Protecting the Network, Securing the Future

Polygon Labs

November 9, 2022

polygon_validator_decentralization|polygon_node_count_hosting_region|polygon_node_count_cloud_provider|polygon_top_validators|polygon_pos_gini_coefficients|polygon_pos_shannon_entropies

Decentralization, or the opposite of having a concentrated locus of control, is among Web3’s core values. It is an ideal state where decision-making power and control over a protocol, project or network are distributed among its participants. One important vector in decentralizing a network are its validators.

On Polygon, a validator is a network participant who locks up MATIC tokens in the system and runs Heimdall validator and Bor block producer nodes in order to help run the network. Validators stake their MATIC tokens as collateral to work for the security of the network and, in exchange for their work, earn rewards. Validators have considerable influence because as block producers they select which transactions will be processed.

There are more than 3.4 billion MATIC worth over $2.78 billion (as of November 2022) locked up in the Polygon validator ecosystem. More than 639 million MATIC of total rewards has been distributed among validators, or over $520 million. You can visit the Polygon validator dashboard for more insights.

Token holders who do not or cannot run a validator node themselves can still participate in securing the network by delegating their stake to validator nodes. Delegators play an important role in the system by choosing which validators to entrust with their tokens. There are a total of about 17,000 delegators, a diverse cohort contributing to the decentralization of the consensus mechanism.

Decentralization can mean different things to different people. For our purposes, a decentralized system would be one where the canonical version of the blockchain doesn’t rely on a concentrated group of actors, but is regulated by off- and on-chain checks and balances to produce desired outcomes.

Here are some factors that can be used to evaluate the decentralization of validators:

• Distribution of validators across geographies and jurisdictions

• Distribution of cloud providers running the validator nodes

• Distribution of staking amongst the validators

• Total distinct entities running the validator nodes

• Governance structure for protocol-related decisions (soft/hard forks)

Let’s see how Polygon stacks up against these metrics.

Geographic Distribution

While the majority of Polygon’s validator nodes are located in either North America or Europe, the distribution across different legal jurisdictions is quite broad. No single country accounts for even a third of the total count.

Cloud Provider Distribution

While two major cloud providers -- Amazon Web Services and Hetzner -- power more than half of validator nodes, neither of them accounts for as much as a third of the total. The remainder is distributed across a long tail of smaller cloud platforms.

Distribution of Staking

The diversity of the validator set can also be seen in the breakdown of who performs the actual block production. Over a two-week period spanning more than half a million blocks, the top-three validators accounted for roughly 10% of the total each.

polygon_top_validators — Source: Polygonscan

Total Distinct Entities

Polygon has a broad group of validators that helps avoid collusion. The number of validators is capped at 100. That list includes digital sports entertainment and gaming giant DraftKings, Coinbase Cloud and Blockdaemon.

There are a number of ways to quantify decentralization. We measured the Gini coefficient and Shannon entropy using publicly available data (see this research paper as reference on methodology).

The Gini coefficient can be used to indicate the inequality of the distribution of validating power among validators. It ranges between 0 and 1 with 1 being most equal. In the blockchain context, a higher Gini coefficient means more validators need to collude to compromise the network and corresponds to a higher degree of decentralization.

The chart below shows the improvement of the Gini coefficient for the Polygon network over time for both fixed and sliding window periods. As reference, Ethereum network’s Gini coefficient is around 0.85.

Shannon entropy measures the degree of randomness and disorder of the distribution of the amount of blocks validated by validators. Intuitively, a higher Shannon entropy means a higher degree of randomness of the distribution of validating power, thus indicating a higher degree of decentralization. The chart below illustrates the improvement in Shannon entropy for the Polygon network. Ethereum network’s Shannon entropy is around 3.5.

Governance Structure

While Polygon launched with a diverse validator set, the network is not entirely permissionless and pretty much the same set of validators has been in place since inception. Enter Polygon Improvement Proposal (PIP).

The PIP process takes inspiration from the PEP scheme, which has become a standard for open source projects, including Ethereum and Bitcoin. PIP-1 outlined a preliminary process for allowing the community to put forward protocol upgrades that improve the network. All changes should originate from the Polygon community forum in the form of a proposal written in line with PIP-1.

PIP-4 outlined steps to enhance the validators’ ability to self-regulate network participation, with the end state being the community assuming responsibility for the network. PIP-4 calls for an increase in the number of validators to 105, a step forward along the decentralization gradient. The proposal recently passed.

A validator can be in one of three states:

Active: The validator is in the current validator set, produces blocks at the Bor layer, participates in Heimdall consensus and commits checkpoint transactions to the Ethereum mainnet.
Notice: The validator sends a transaction to unbond. Before entering into the unbonding period, the validator needs to be in active state creating, signing and proposing blocks for a certain time.
Unbonding: The validator is inactive in this state and thus earns no reward.

PIP-4 introduce the following roadmap for admission and onboarding of new validators into the set:

Phase 1 (current) - The multi-sig holders would set the admission parameters and choose the validators.
Phase 2 (next step) - The governance and validator communities set and approve the parameters, and the multi-sig holders implement those parameters.
Phase 3 (end state) - The governance and validator communities set the parameters and admit and onboard validators themselves.

The Polygon PoS consensus engine (Heimdall) has an inbuilt governance module that can synchronously carry out consensus parameter changes across the network. Proposals can be submitted to the module along with a deposit containing the proposed changes. Votes cast by validators are tallied by each validator, once the defined voting parameters are met each validator automatically updates itself with the proposal data.

The current voting parameters (denominated in staked MATIC):

Quorum: 33.4%
Threshold: 50%
Veto: 33.4%

A list of the changeable parameters by the Heimdall governance module are available here.

In the next post, we will take a deep dive into the MATIC token itself, its various uses, distribution history and what that means for decentralization.

Head over to support channels for validator and staking queries for more information, join the governance conversation on our forum to suggest ideas and tune in to our blog for the latest on Polygon.

Let’s make decentralized Web a reality!