As the Ethereum staking ecosystem continues to evolve, understanding validator performance becomes increasingly crucial for both seasoned stakers and newcomers alike. At Figment, we recognize that grasping performance metrics requires a holistic understanding of the Ethereum network.
Before diving into this topic, we strongly recommend reading our previous articles in this series: “Beyond the Basics: “Understanding Rewards on Ethereum” and “Beyond the Basics: Navigating Ethereum Penalties.” These pieces lay the essential groundwork for comprehending the full spectrum of Ethereum staking economics.
Building on that foundation, this article takes you deeper into the intricacies of assessing validator performance on Ethereum. We’ll explore why there’s no simple substitute for a comprehensive understanding of the network, dissect common performance metrics, and reveal the hidden complexities behind seemingly straightforward calculations. From APR misconceptions to the nuances of ‘all-in-one’ metrics, we’ll guide you through the potential pitfalls of performance assessment and illuminate the often-overlooked aspects that don’t appear on-chain but are crucial for long-term success.
Assessing Performance on Ethereum
Simply put, there is no substitute for understanding Ethereum when assessing performance. You can find all kinds of metrics that purport to help understand a validator’s or group of validators’ performance. These metrics can be useful but caution should be heeded in using them. When using any metric there are a number of things to keep in mind. Let’s take a look further.
For instance, what metrics accurately reflect what it’s intended to measure?
One simple example is Annual Percentage Rate, or APR. A common approach to calculating the APR for a given validator, or group of validators, is to measure rewards and balance over a given period, say 30 days, and then annualize this by extrapolating the same reward over the next 12 months.
However, since the launch of Beacon Chain at the end of 2020, the number of validators has been increasing.
(source)
This means that in all but a small number of cases, this approach would have overestimated the consensus layer (CL) reward rate as the per validator reward rate decreases with an increasing number of validators. A similar dynamic exists for execution layer (EL) rewards, but these rewards are more volatile so the story isn’t as cut and dry here.
Understanding Validator Performance: Beyond Simple Metrics
The best example in this case is missed attestations. The metric is simple enough – how many attestations did a validator fail to earn rewards for of the 225 slots per day they were expected to attest? The metric is simple, but the conclusion drawn is anything but. Is poor performance on this metric a problem with a piece, or pieces, of the validator’s infrastructure? Or was the validator offline during a given period due to maintenance? Or, is it that the network as a whole is experiencing difficulty? For a validator to earn attestation rewards on Ethereum, another validator must include that attestation in a block – so sometimes missed attestations are a result of other validators failing to perform their duties.
Using ‘all-in-one’ metrics is another interesting case. The idea behind an ‘all-in-one’ metric is that it takes in data related to most, or all, of a validator’s duties and computes a score. A user of the metric can then, supposedly, make a judgment about the validator’s performance.
One of the issues in constructing such a metric is choosing the importance of each validator duty; in other words, how do you weigh each component of the metric? An intuitive approach is to use the weighting of the protocol (see the pie chart above). The problem is that the weighting is relevant for performant validators but less accurate for non-performant ones. The reason is that failing to attest to the target and source checkpoints carries penalties. So for a non-performant validator, the weights on these pieces should be higher.
Beyond this issue is the fact that rewards change over time. So if the metric you are using appears to measure performance, it must accurately take into account how rewards change over time as more validators are added (or removed) from the active set. As already mentioned, the per-validator CL reward rate decreases as the number of validators increases over time. EL rewards have a similar dynamic but change at a different rate – so the relative weight of CL to EL rewards changes over time as well.
Data set issues are another key consideration when using metrics and evaluating performance. This is especially relevant when considering the performance of pools or staking service providers.
Some specific issues related to data sets:
- Does the data set capture all (or most) of an entity’s validators?
- How is survivorship bias handled?
- Validators that were once part of a pool but were slashed have likely been removed – how has this been incorporated in the metric?
- Average vs median – especially relevant during volatile periods, EL rewards can be very positively skewed, resulting in a wedge between the average and median
- This is especially important for customers of staking service providers /pools to understand. Often the average performance is posted, but the typical, i.e., median, customer/user will experience a lower reward rate thanks to skewness
It is just as important to appreciate that metrics can only take you so far. Some of the most important aspects of running a validator on Ethereum are things that aren’t captured on-chain.
Some specific examples:
- Is slashing protection being used?
- What precautions is your staking service provider taking concerning validator key security?
- Does the staking service provider offer coverage?
- What happens in a disaster scenario? How will a user/customer be able to access their stake?
- What is a staking service provider’s stance on compliance?
These are just a few examples of important issues. These items speak more to risks and point to the obvious fact that reward maximization should not take priority over risk minimization. Earning the highest possible reward rate is useless if a staking service provider is taking risks with a customer’s stake.
How is Figment Different?
Figment emphasizes risk-adjusted rewards in our staking operations. Our approach carefully balances reward optimization with comprehensive risk management practices. We achieve this through:
- Multiple layers of slashing protection and monitoring
- Enterprise-grade key management systems with strict security protocols
- Regular security audits and disaster recovery testing
- Clear procedures for stake withdrawal and emergency scenarios
- Full regulatory compliance and transparent reporting
By focusing on risk-adjusted rewards, we ensure our clients can maximize their staking rewards while maintaining the highest security standards for their assets. This measured approach has proven especially valuable during network upgrades and periods of market volatility, where our risk management framework has helped protect client stakes while maintaining consistent rewards.
How Does Figment Think About Performance?
Safety and security of stake are of primary importance. At Figment, our risk-adjusted rewards framework ensures that performance evaluation always prioritizes the protection of our clients’ assets while optimizing rewards.
Validator performance consists of two main components: Consensus Layer (CL) and Execution Layer (EL) rewards. CL rewards serve as a key indicator of operational excellence and network health. These rewards come primarily from attestations – validators must perform 225 daily attestations with precise timing and accuracy. While some performance factors may be outside an operator’s control due to network conditions, Figment’s infrastructure is optimized to maximize attestation effectiveness within our risk parameters.
EL rewards, however, stem from block building decisions and statistical probability. With over 90% of blocks now utilizing MEV-Boost, these rewards largely depend on relay selection and timing strategies. Luck plays a significant role – both in how frequently a validator is chosen to propose blocks and the timing of those proposals relative to network activity.
Figment applies sophisticated analysis to differentiate between luck-based outcomes and genuine outperformance, allowing us to make informed decisions that optimize risk-adjusted rewards. This data-driven approach, combined with our emphasis on security and stability, enables us to deliver consistent rewards while maintaining the highest standards of asset protection.
Conclusion
Understanding Ethereum staking performance requires embracing its complexity while maintaining perspective on what truly matters. While abundant metrics and data are available, the key is focusing on meaningful indicators that align with a comprehensive risk-adjusted rewards strategy. Instead of chasing the highest possible rewards, successful staking requires a balanced approach that considers both quantifiable metrics and crucial off-chain factors.
At Figment, we recognize that many critical aspects of validator performance—such as security protocols, disaster recovery procedures, and risk management frameworks—aren’t captured in on-chain metrics. Our approach to risk-adjusted rewards combines sophisticated data analysis with enterprise-grade security measures and operational excellence. As Ethereum continues to evolve, we adapt our assessment methods while maintaining our unwavering commitment to protecting our clients’ stakes.
Ready to optimize your Ethereum staking performance without compromising on security? Contact Figment today.
