Translation: Blockchain in Vernacular
We surveyed independent operators (often broadly referred to as “independent stakers”) to understand their profiles, demographics, concerns, and motivations to provide better insights. Despite widespread feelings of structural deprivation and concerns about validator centralization among survey respondents, we noticed that their confidence and resilience in staking was very high.
The data is intended to provide perspective in the words of these privacy-oriented participants themselves so that their needs are accurately represented. This survey is intended to be conducted annually, and feedback on the question set is welcome.
1. Methods
1) Collection and distribution
The survey was collected using LimeSurvey software. Questions used branching display logic to ensure continued relevance to the respondent. Cookies were used to prevent duplicate participation and CAPTCHAs were used to prevent bot activity. The survey was open to the public and the identities of the respondents were kept anonymous.
We solicited responses on EthStaker’s social channels (Reddit, Discord, Twitter, Farcaster), Obol’s Twitter account, and the public channels of the most popular staking service providers, hardware providers, and client software. The survey was also published on the Beaconcha.in website, Rhino Review, and the Week in Ethereum newsletter. Submissions were collected from April 8, 2024 to May 6, 2024.
analyze
Only complete survey data were used. Incomplete surveys were discarded. Results were manually checked and no complete survey data was discarded. The data shown in the pie chart is the result of the single-choice questions. Multiple-choice questions are represented in the discussion by the asterisk ✶ Unicode ✶ character ✶.
2) Sampling bias
While the data shows a preference for home-based staking over more detached staking methods, this may be related to the identity of the participants and who the survey was able to reach. Stakers who frequently participate in the staking community tend to be those who manage their configurations. Those who do not manage their own configurations can choose to ignore it, as they generally do not need immediate attention to maintain their validator nodes.
This information is more reliable qualitatively than quantitatively because it relies on subjective data from a self-selected subset of stakers.
3) How widespread is this data?
Publicly available node crawlers put the number of Ethereum nodes at between 6,000 and 11,000. Not all of these are validators, many are managed by professional operators. This survey is only for stakers who stake with their own capital, questions are not relevant to professional operators. As of this writing, the number of validators using Rocket Pool can be estimated to be 1,832, based on the number of nodes with ETH in the staking snapshot, and subtracting the number of Allnodes nodes. This can be used as a rough lower bound for the number of independent operators. Of the 1,024 total responses, 868 were from stakers who claimed to control their node configuration. We can estimate that between 8% and 47% of node operators responded to this survey, but the likelihood is lower. Keep in mind that this is a percentage of all node operators, which includes professional operators. This survey focuses on non-professional operators.
While it is easy to see the number of validators on the network, there is currently no way to accurately count the number of validating nodes, individual operators, or even just the number of nodes on the network (this can be considered a feature rather than a bug). Node operators can choose to self-identify their validators, but most independent operators and many professional operators do not do so.
2. Results
The raw data can be found here https://github.com/eth-educators/staking-survey-data.
1) Overview of the respondents
32% are genesis stakers.
80% are home-based stakers, with an additional 4% pledging both at home and remotely.
84% do not hold any significant amount of liquid staked tokens.
85% did not change their primary staking method.
77% of people use between 66% and 100% of their Ethereum in staking.
30% use smooth staking pools, 61% do not, and 9% are unfamiliar with the concept or are unaware of the options available to them.
95% use Linux to run their validator(s)✶.
51% did not use staking software to set up their validator. 27% used Rocket Pool, 15% used DAppNode, and 10% used Eth Docker (“did not use staking software” usually means they followed the guide and used systemd) ✶.
85% have not changed their staking method since they started staking.
Figure 1: When did you first start running a validator?
Figure 2: Where do you stake, do you hold LST (Liquid Staked Tokens)?
Figure 3: Have you changed the way you stake and how much ETH you stake?
Figure 4: Are you participating in the smoothing pool?
Figure 5: What operating system are you using for staking?
Figure 6: Do you use staking assistance software?
2) Main focus
Key concerns include super majority client risk, suboptimal tax structures for staking, hardware issues, and key management.
69% of respondents do not track their bandwidth usage, and 78% are unsure if it has increased since Dencun.
Respondents spend an average of 3.4 hours per month (median 2 hours) maintaining their setup. Excluding one outlier data point with 22 standard deviations (x = 155 hours), the average time is 3.2 hours per month.
On a 1-10 rating scale, with 10 being the greatest possible risk to the network, stakers rated supermajority risk at 7.4 and stake concentration risk at 7.2 (the ratings have been adjusted for clarity).
When asked what percentage of Ethereum they expect to be staked in the next 2-3 years, the average response (n = 1003) was 49.4%.
When asked what percentage of staked ETH they guessed was operated by independent operators like themselves, the average response (n = 924) was 15.9%.
Figure 7: What do you care about most?
Figure 8: How much bandwidth does your node use?
Figure 9: How many hours per month do you spend on maintenance?
Figure 10: What are the risk factors for the network?
3) Value preservation and representativeness
89% of respondents believe independent stakers are either more or equally important to the network than when they first started staking (11% believe they are less valuable).
66% of respondents believe that independent stakers receive equal or greater benefits from participating in the consensus process than they did when they first started staking (34% believe they receive less value).
When asked how well they felt their interests were represented in ongoing research and protocol development, the average score was 5.8 out of 10, with 1 being “very well represented” and 1 being “not represented at all” (the scores were adjusted for clarity).
50% of respondents believe that protocol research either ignores independent stakers or is powerless in the face of wealthy interest groups (19% report that they are not paying attention).
92% of respondents supported or were neutral that the issuance curve needs to change to better incentivize decentralized forms of staking (regardless of existing proposals).
Figure 11: What is the value of independent staking?
Figure 12: What does circulation and representation in research mean?
Figure 13: Advocacy for independent stakers
4) Continue to participate
The primary motivation for staking initially was to support the Ethereum protocol (84%), followed by earning a yield (81%).
65% plan to continue adding new stakes.
35% plan to continue adding new stakes, but intend to stop when certain external conditions are met.
31% have no plans to add new stakes.
62% said they had no plans to quit
Figure 14: Do you plan to add new stakes?
Figure 15: How long do you expect to continue running a validator?
5) Learn where stakers come from
69% of respondents cited technical guides as their primary source of learning (e.g. Someresat, CoinCashew, Rocket Pool documentation, client documentation).
63% of respondents cited EthStaker as their primary source of learning.
53% of respondents cited ethereum.org as their primary source of learning.
In terms of following up on necessary updates and protocol research, the most commonly reported sources were Discord, blockchain explorer notification services, Twitter, Reddit, ethresear.ch, and podcasts, with a large portion specifically mentioning The Daily Gwei.
Figure 16: What sources did you use to learn about staking?
Figure 17: What sources do you use to follow up on protocol research?
6) Open Questions: Unresolved Concerns
At the end of the survey, stakers had the opportunity to comment on any questions that were not fully covered in the survey. The full responses are available in the raw data, and an AI-assisted summary of the responses is provided here:
Decentralization and Centralization Risks: Many respondents are concerned about the centralizing effects of current protocol development and the underrepresentation of independent stakers. Liquid Staking Tokens (LST) are seen as a centralizing force that makes independent node operations less attractive and viable, and re-staking is seen as a potential factor for centralization.
User experience, technical barriers, and hardware: There were many calls to simplify the staking process and make it more accessible to non-technical users. Issues with bandwidth, the need for IPV6 support, and the need for asset planning solutions were also mentioned.
Emission Curve Adjustment: There are significant concerns about the impact of changing the issuance curve, especially the potential impact on independent operators, and some support this research direction. Many believe that reducing returns will result in only large centralized staking entities being able to obtain favorable conditions.
Economic and Tax Impact: Tax policy is a significant burden for independent stakers, making independent staking less economically viable than holding Liquid Staking Tokens (LST), which enjoy more favorable tax treatment.
MEV: Opinions on the use of MEV (Maximize Transaction Value) are split, with some choosing not to enable it due to ethical concerns. These respondents advocated for finding alternative solutions and conducting further research to address the issues posed by MEV. MEV is seen as a potential tool that could enable large centralized entities to dominate the protocol. Respondents called for stronger measures to prevent MEV from undermining the decentralized nature of Ethereum.
Privacy concerns: Stakers are concerned about the disclosure of details of their operations, such as IP addresses and transaction inclusion, which compromises their privacy and security. Respondents hope to develop and integrate privacy-preserving technologies. Privacy concerns also touch on the issue of censorship, with respondents noting that a lack of privacy would force some stakers to censor transactions, undermining the neutrality and inclusiveness of the network.
7) Demographic Information
95% self-reported as male.
90% believe they have some or strong technical abilities.
74% do not work in cryptocurrency.
88% of staking comes from North America, Europe or Australia.
Figure 18: Technical background and employment industry
Figure 19: Which region are you staking from?
Independent operators are predominantly skilled males from North America, Europe, and Australia, using Linux operating systems, and these results are not surprising — these imbalances have multiple causes and have been the target of many diversity initiatives in the staking community. However, less is known about the operators’ staked percentages, and what percentage of ETH they use to run validators, as stakers tend to be reluctant to share information that could reveal vulnerabilities to their security.
The large number of actively engaged Genesis stakers, with 80% running from home, 84% not holding any significant amount of liquid staked tokens, and 77% staking more than 66% of their total ETH, are new and encouraging data that demonstrates a high level of confidence and resilience among independent operators in staking.
8) Proportion of independent operators
A recent report from StakeCat looked at addresses identified as independent operators (methodology here) and determined that the percentage of independent operators on the network has increased since the merge. This survey data shows that a significant portion of respondents (32%) have been staking since the EthStaker Genesis event. Since the survey data was collected from a voluntary subset of stakers, it is likely that these findings are biased towards stakers who joined during the Genesis period, when EthStaker was the only comprehensive source of staking education and support.
Now that independent operators are increasingly joining through protocols and products with improved user experience and their own support, these subsequent groups of stakers may be less connected to the overall staking community. This is desirable because software, education, and support for independent operators is shifting toward a diverse set of self-sustainable projects that may attract a less “tech-savvy” set of operators.
Over time, the amount of ETH staked has increased since the Shapella hard fork enabled withdrawals, and at the same time, there has been interest from professional entities that utilize delegated staking.
Figure 20: The amount of ETH staked shows an increasing trend over time: after the Shapella hard fork, the slope increased significantly.
These specialized entities often benefit from economies of scale, as they can run hundreds or even thousands of validators per node, reducing hardware costs per staked ETH, while independent operators typically only run a few or dozens of validators per node. Large entities can also pool execution layer rewards together to balance their validator portfolio, an option that has only recently become available to independent operators with the introduction of two independent staker balancing pools. Specialized operators can also earn additional benefits by participating in extra-protocol services that require more powerful hardware or operational expertise. These factors result in sometimes higher rewards for LST holders than independent operators, even after taking into account fees paid to staking service providers.
Survey respondents expressed feelings of disenfranchisement when considering these results. They feel their role in the network is just as important or more important than when they first started staking, but do not feel that the protocol values their participation in the same way. When asked about support for representation of their interests in protocol research, the majority were neutral between “no representation” and “very good representation,” leaning slightly toward the latter. Over half of respondents felt that research was either hostile, negligent, or ineffective in representing their interests. The majority expressed support for changing the protocol’s reward structure to address this feeling of disenfranchisement (understand that the above support does not necessarily indicate endorsement of existing proposals).
9) The “stickiness” of independent operators
Over the years, independent operators have been called “irrational actors” and “altruists,” and their stakes have been argued to be more “sticky” than those injected by delegated stakers. It’s not possible to directly compare the two groups from this survey, but respondents did express their desire to continue staking, regardless of small fluctuations in the annualized yield or changes in the staking ecosystem as a whole.
While the majority of respondents expect to continue staking at any level of positive returns, 21% said they have a specific return threshold below which they would exit staking, with the average value for this threshold being 2.3%*.
(*It’s worth noting that this number should be understood to be biased by a legitimate desire to keep revenue high, as some stakers have indicated that they are currently running validators but would unstake at a threshold we’ve long exceeded, suggesting that some may not be monitoring actual current revenue consistently.)
This reluctance to unstake, combined with the ideological motivation for staking in the first place (“to support the Ethereum protocol”), is often attributed to independent operators’ feelings of “altruism”. It is important to note that their inertia can be bidirectional. Independent operators are unlikely to unstake their stake in response to small fluctuations in the market or issuance, but are equally unlikely to re-stake it to the network after it has been removed, even if conditions are slightly favorable. As a result, their losses may be more persistent than those of delegated stakers. Long-term reliance on altruistic motivations rather than structural fairness for independent operators may gradually eliminate this portion of stakers.
The series of airdrops to independent operators in 2023 and 2024 recognizes the disparity in rewards and power between independent operators and professional operators, and the value of keeping independent operators on the network. These one-time incentives are helpful for the transition period, but they should not be relied upon to close the gap in the long term.
3. Looking to the future
StakeCat's recent report shows that demand for independent staking remains strong, and the ratio of independent operators to professional operators has remained relatively stable over time, or even increased. Much of the recent research aims to protect the meaningful participation of independent operators amid the centralization pressures emerging from current protocol designs. Such proposals should rely heavily on insights gained directly from the motivations and concerns of independent operators, and this survey and report are intended to provide this data for brainstorming and research purposes only.