Algorand Decentralization Dashboard (xGov Proposal 3264792007)

Governance Discussions xGov Proposals
1

Description

Valar Solutions GmbH has developed the Algorand Decentralization Dashboard, which provides an easy-to-understand analysis of Algorand’s decentralization. It analyzes the network’s online stake in real time. This includes native staking as well as third-party staking solutions like Valar peer-to-peer staking, Réti pools, and liquid staking tokens (LSTs) of Tinyman (tALGO) and Folks Finance (xALGO).

Users can explore and gain real-time insights into:

  • who controls the stake,
  • who is operating the stake,
  • through which solutions the stake is being operated,
  • how many users different staking solutions and node operators have, and
  • what are Algorand’s metrics for liveliness, collusion-resistance, and upgradability.

This enables users to make well-informed decisions when selecting a staking solution or a node operator, thereby contributing to increased network decentralization.

To the best of our knowledge, this is currently the most detailed, transparent, and real-time analysis of Algorand’s online stake decentralization. The project can be run locally. The majority of data is fetched directly from the blockchain, i.e. there is no dedicated backend or trust needed. Some third-party APIs from Nodely are used to speed up data fetching. The project includes one of the most complete, publicly accessible lists of publicly doxxed Algorand accounts.

A deployment of the dashboard can be accessed at: https://decentralization.valar.solutions/

The project is open-sourced: GitHub - ValarStaking/algorand-decentralization: Algorand Decentralization: Online Stake Analysis

About the Team

Valar Solutions GmbH was founded by Dr. Uroš Hudomalj and Dr. Alexander Marinšek after winning Algorand’s Global Hackathon - Change The Game, at Decipher 2024, with the IgoProtect staking platform. The platform evolved into Valar peer-to-peer staking, which achieved a peak of 100M ALGO staked by more than 350 users staking via 100 different node runners, who operate more than 15% of all online accounts.

In addition, the company has developed:

  • xGov integration for Réti that enables stakers to express their opinions about xGov proposals and validators to easily vote based on their wishes.
  • VERA platform for transparent, fair and verifiable random draws.
  • Certify platform for on-chain credentials for node runners, which won 1st place at Algorand France Hackathon.
  • Block Anxiety - a tool helping to assess your staking performance, which has been integrated e.g. into https://algonoderewards.com/.
  • Block Tracker - a tool for tracking block producers on-chain.

Valar Solutions GmbH is also offering node running services via Valar’s peer-to-peer staking platform and Réti staking pools, and has contributed to Folks Finance audit competitions for their xALGO liquid staking token (1st place) and Folks Smart Contract Library (4th place).

Dr. Uroš Hudomalj currently serves as a member of the xGov Council. He will abstain from voting on this proposal within the xGov Council to prevent conflict of interest.

Additional info

This effort is in line with Valar’s mission to increase decentralization of Algorand. To achieve this goal, it is necessary for the ecosystem and its users to have transparent and objective metrics to evaluate decentralization. This project aims to be the basis for this by transparently aggregating information that is typically scattered among different sources, opening doors for efficient joint collaboration within the ecosystem.

Adoption Metrics

Social media contributions based on this project have generated on X, Reddit, LinkedIn, Discord, and Telegram:

  • over 30k impressions within 1 month
  • over 1.5k engagements within 1 month
  • over 500 likes within 1 month
  • over 100 reposts within 1 month

This proposal requests 24,999 ALGO in retroactive funding to cover the completed development.

Link to proposal: xGov

Happy to answer any questions and hear any feedback on how to improve the Dashboard further!

9 Likes
2

Valar is fantastic platform. The team have always been incredibly helpful and responsive to all feedback and support. Their contribution has greatly improved the decentralisation and resiliency of the network while having the ability to maintain assets in the delegators accounts.

Their products are in high demand and improve visibility for new comers which is beneficial to the ecosystem.

The Nakamoto Coefficient metric is interesting.

I wasn’t aware 2 entities working together could attack the chain, how is that determined?

2 Likes
3

Thanks for the kind words!

About the Nakamoto Coefficient, it is typeically defined as the minimum number of independent entities that would need to collude to attack the chain, i.e. historically this meant for Bitcoin to do a 51% attack. On Algorand and other PoS networks, which have a lower threshold of 33%, that limit is used for the online stake.

image
image303×257 13.5 KB

Currently, on Algorand there are a few very large operators - besides the Algorand Foundation (+20%), also Kiln (+20%) - which has onboarded majority of exchanges, and Folks Finance (~10% + 5% delegated to Defly - in the Dashboard considered as a separate entity), which could collude and reach more than 33% of online stake.

image
image969×456 75.5 KB

3 Likes
4

Amazing work by Valar Solutions. I support the proposal. Thank you Valar team.

2 Likes
5

Thank you for your support!

1 Like
6

There is no 33% threshold on Algorand though. For Algorand, if you consider 51% to be the threshold for Bitcoin, then its the threshold to control the chain, not hinder/halt it. To control Algorand you could look at soft and certification vote thresholds - 75.8% and 74.1% ( Data Types - Algorand Specifications )
I think for the NC I calculate I’m conservative and just use the 74% threshold. By accounts at least, this yields 41.
If you want to look at % to simply halt the chain then it’s only around 24% but that wouldn’t coincide with 51% as bitcoin equivalent.

1 Like
7

Thanks for the comment!

The 33% threshold was chosen as per the initial papers about the Algorand protocol [1,2]. The value is also referred to by the Algorand Foundation [3]. Moreover, the same limit is typically selected for the Nakamoto Coefficient with other PoS blockchain networks, which makes comparison simpler. A newer work by Fabrice Benhamouda, et al. [4] interestingly notes that the Algorand protocol is designed to tolerate 20% of adversaries, proving protocol ensures consistency with a certain probability under this condition. They also attempt to prove protocol liveliness under different network conditions, which are larger and dependent on the network conditions. The Decentralization Dashboard separately shows a liveliness coefficient, where the limit of 20% stake is used for simplicity. As the dashboard is open-source, you can easily replace the displayed limits by changing the used constants.

image
image305×314 16.4 KB

References:
[1] J. Chen and S. Micali, Algorand: A secure and efficient distributed ledger, Theoretical Computer Science, 2019, DOI: 10.1016/j.tcs.2019.02.001
[2] J. Chen, S. Gorbunov, S. Micali, and G. Vlachos, ALGORAND AGREEMENT: Super Fast and Partition Resilient Byzantine Agreement, 2018, Accessed on: 2025-10-26, Available at: https://eprint.iacr.org/2018/377.pdf
[3] Algorand Foundation, Reaching consensus through Pure Proof-of-Stake, 2025, Accessed on: 2025-10-26, Available at: Pure Proof-of-Stake - how Algorand's consensus mechanism works
[4] F. Benhamouda, et al., Analyzing the Real-World Security of the Algorand Blockchain, 2023, Accessed on: 2025-10-26, Available at: Analyzing the Real-World Security of the Algorand Blockchain

1 Like
8

The original algorand paper also talked about forking and didn’t have the soft vote / certification vote steps so frankly can and should be, ignored. It’s definitely been a source of consternation over the years that it has’t been replaced/updated and people have instead had to just point to the specs repo.

Fabrice talking about 20% is likely around that being the approximate threshold that can halt the chain (which is certainly an important threshold !). For full control, proposing a block and guaranteeing it was accepted, you’d need 74+% of all online stake (and across multiple accounts you controlled). With the comparison to Bitcoin - the 74+% threshold makes more sense (IMO at least) for the NC of Algorand.

3 Likes