This week, Blockdaemon has launched several new updates, but we wanted to highlight one which we feel especially strongly about: the decentralization score, which is shown on the dashboard for all protocols we support (Bitcoin, Ethereum, Aion, GoChain, Quorum, Fabric, today, with Algorand coming soon).
Going forward, this decentralization score is going to be a key component of monitoring and understanding the health of your network, public or private. Of course, what we’ve launched is a simple first version. The algorithm is going to grow and evolve over time as we collaborate with partners while continuing to add meaningful ways to measure key attributes. We’ll also be exploring with customers and partners ways to dynamically distribute nodes to maximize decentralization, as well as setting thresholds and additional rules based on the size of the network.
It’s easy to dismiss decentralization and consider it a disposable aspect of blockchain technology — “wouldn’t this thing work faster if we centralized aspects of it” — but it’s the very heart and soul. When you examine real-world use-cases where blockchains bring actual value, decentralization is a hard requirement for accountability and validity of data. Otherwise, you’ve succeeded at implementing a slow and inefficient database. In any case where auditing is a possibility — multi-party finance, regulatory compliance and so on — decentralization is the primary check against multi-party collusion.
Our decentralization score is currently optimized for private Ethereum networks, but we will be improving it for other networks over time. We measure five attributes right now:
We want to see a healthy network existing in multiple clouds. It’s no coincidence that we’ve allowed deployment into multiple clouds from day one — if there is a core tenant for Blockdaemon, “multi-cloud” is close to it.
The current public clouds are fantastic feats of technology and engineering. They have certainly ushered in a new era for development and lowered barriers to entry. However, it’s an undeniable risk of centralization.
This has also been a focus for Blockdaemon since our earliest efforts. We don’t differentiate between geographic zones, provided that the availability zone for that cloud provider allows us for reliable node launches (not all do). We believe networks should transcend borders and as appropriate, not be at the risk of any single nation’s regulatory decision.
We have seen heavy activity among our user base outside the expected corridor of North America, Europe and Eastern Asia. We want to ensure that these users are able to place nodes in areas that have low latency and perhaps favorable regulations.
Beyond that, cryptocurrencies have especially been concerned with regulatory capture in the Western world, and want to be able to exist above and beyond those borders. For all these reasons, we consider this an important aspect of the score.
This is the first order of decentralization most think about in terms of blockchains — an armada of nodes under the control of one entity is a threat to the viability of the network. We ensure that when we are measuring this, we are counting distinctly identified and billed users. We don’t count multiple subscribers to a shared node as impacting the decentralization meaningfully.
This measure is exceedingly important, and we will be looking for better ways to understand ownership on public networks. Given the community’s values of anonymity, this is not straightforward, so we measure what we can verify currently.
If ownership highlights the risk of one owner of many nodes, a low node count is the other side of the coin. A small network is vulnerable to malicious actors. A small network is at increased risk of disruption, as there is not a robust web of connections linking the nodes together.
We also see node count as a measure of health for that network. It’s no surprise that the networks with the most name recognition also carry the highest node count. When coupled with the Ownership axis above, we know that many owners with many nodes is sign that the project has gained value and acceptance outside of a core.
What are “External Nodes”? They’re nodes outside of our network. Blockdaemon doesn’t pretend to be the only solution for launching nodes — nor should we be the only tool for deployment in a healthy and thriving ecosystem. Though we believe we make it easy to orchestrate cross-cloud deployments worldwide, a decentralized network cannot rely solely on Blockdaemon. We’ve heard this critique or Blockdaemon before and this is our way of acknowledging and agreeing with it.
The Road Ahead
We know that our first effort is just a toe in the water, and already have seen additional suggestions for augmenting these measures, or measuring via other heuristics based on network size. We also know there are other higher level concepts that should be reflected in certain cases — hash rate control, centralization of assets, etc — which we will be finding ways to accurately layer in.
Decentralization is a huge topic and it tends to be a genie that won’t go back in its bottle once its power is realized. It’s imperative that as a community we remain mindful and honest in our assessments of our efforts to decentralize, even when our scores aren’t what we’d hope. While it may be easier to solve problems through centralization, the breakthroughs and accountability afforded by decentralization will outweigh the challenges in these early days of implementation.