This post is to follow up on my rationale for how I distributed my vote for the Optimism RetroPGF Round 2 and some thoughts relating to the results, a retrospective and how RPGF3+ could help better distribute future rounds to public goods.

When it comes to funding public goods, there remains a dilemma of how a Badgeholder should go about defining what a project must have demonstrated or will demonstrate to be allocated some funding from RPGF. These views remain very diverse and experimental, which is great! I think the mix of formulas from each Badgeholder ensures a good diverse set of viewpoints.

This dilemma is whether or not we reward for retroactive impact made up to this current point, or we reward projects with potential future impact in need of funding to offset risks. Perhaps these are two different visions and may constitute two different funds with different objectives. I tried to balance both these views in how I voted, leading to an allocation of greater than 0 to a total of 183 projects, with the highest one at 1.3%.

The goods that generally made it to the top were still projects that have made a proven impact in our ecosystem including BLST by Supranational (1.3%), Keccak hashing (1.2%), Ipsilon (1.2%) and the monorepo dependencies collection (1.3%). These are undoubtedly projects that make the Ethereum/Optimism engine run and rewarding teams that make it run better, including researchers! The highest rated experimental allocation is NiceNode, getting 1.2% of my vote. This was interesting because as a Protocol contributor, I understand the great need for better node running UX in general, but NiceNode is also not a productionized project that has made a sizeable impact… yet (no pressure guys!). These projects should not be sidelined from public goods resources because they are taking great risk to solve a need in the community.

For education, the top allocations went to ETHGlobal at 1% and 0.8% each went to RadicalXChange, L2BEAT, Kernel, IC3, CryptoStats and BuidlGuidl. There were some disadvantages of my allocation methodology that made Education votes (which I’m the least in-tune with) less impactful than I think they should be, which will be discussed below.

As a Protocol contributor, I inherently have biases on what I think is useful. This experiment helped open my eyes to realize how important every layer of the stack is and the awesome projects people are building on top of what we do at Layer 1. To minimize this bias, I allocated 33.3% of my vote to Infrastructure, 33.3% to Tooling and 33.4% to Education.

Upon reviewing the results, the flaw in this methodology is that some categories are more saturated than others and also mislabelling of a category could’ve cost your project some allocation. In fact, some can even argue that a project is Infrastructure instead of Tooling, which would’ve made an allocation difference in my methodology. Due to the lack of Infrastructure projects comparatively to Tooling and Education, every infrastructure project got a greater share of allocation just because it was in a less crowded category. Perhaps a better strategy would’ve been to allocate by category ratios.

Additionally, Education is such a hard thing to evaluate in terms of quantity and quality. It also requires extensive research as to what your project/DAO/community has accomplished over the years and it’s hard to evaluate 73 of these equitably in a short period of time. I’ve recommended projects (alongside other badgeholders) to dedicate a one-pager, tweet thread or some other quick medium of explaining what they’ve developed, done or will do to educate users of our ecosystems. Many did not and those who did had a huge advantage in getting votes from me. I generally evaluated the content, quality, tools, target audience and other quantifiable metrics to gauge impact. In essence, if you had one medium of distribution (e.g. videos only on YouTube), you had less allocation than someone who put together an entire course with templates in a repository alongside their videos. Anything unique that you did scored you bonus points. I’m always interested in innovative ways to onboard new users into the ecosystem.

I made a huge effort to dig deeper in the education category because I believe it’s important and where I’m most ignorant. With tooling and infrastructure, it’s pretty straight forward what the problem is and how they’re trying to solve it with their code/product. For education, especially if courses and newsletters are mixed in with DAOs and NFT projects, makes it hard to distinguish the noise from the signal.

Part of what I believe we are solving for here is to provide resources to projects that are offering a good/service to everyone at a loss, in order to fulfill a demand in the community that isn’t expected to directly generate revenue. I think this requirement is what distinguishes a public good from a (pre)seed/angel investment. So when making allocations, I feel that:

• The most important criteria is that the good is useful or it may be useful in the future for our ecosystem. The more parts of the ecosystem it benefits, the better.

• Secondly, the usage of the good is free and will stay free for everyone. Therefore, somebody or some entity is taking a loss for running and maintaining this good.

• Thirdly, if you’re backed by VC funding or a for-profit company/product, I feel that the needs impact isn’t as great as you have access to resources at your disposal or have the ability to raise more resources.

These beliefs in how I define a truly public good may have affected the individual project allocations for RPGF.

Badgeholders are generally very active members in the Ethereum ecosystem and we have a limited amount of time to dedicate for research and due diligence (we do this voluntarily!). During the process, it was brought to my attention that there were some projects that potentially misused previous grant funds or had violated codes of conduct as the voting period was ongoing. Some badgeholders had even submitted votes before allegations were brought to our attention and I think we can all agree allegations take time to resolve. There should be some formality in how badgeholders should hastily receive these disclosures, so we can account for this in our voting. At the same time, we need to prevent the system from being abused to nefariously accuse other projects for violations with little to no merit in their allegations. I, and potentially other badgeholders, would have disqualified projects undergoing some vote, investigation or accusations with merit until matters were resolved. RetroPGF does intend to retroactively reward projects in future rounds, so missing out on a specific round until matters are resolved seems to be a prudent way to handle these governance issues.

I highly recommend that you look at other badgeholders and how they’ve allocated their votes so we can build an optimal, effective and fair way of distributing resources to public goods in need of resources. I definitely spent more time on this than I would next round, so I would need to optimize my strategy going forward. Optimism’s voting rationale post, alongside the Twitters and blog posts by other badgeholders are a great way to expand the social experiment. I hope we can figure out how to better shape RPGF3+ together and create a vibrant and sustainable public goods ecosystem for the future!

This post is to follow up on my rationale for how I distributed my vote for the Optimism RetroPGF Round 2 and some thoughts relating to the results, a retrospective and how RPGF3+ could help better distribute future rounds to public goods.

When it comes to funding public goods, there remains a dilemma of how a Badgeholder should go about defining what a project must have demonstrated or will demonstrate to be allocated some funding from RPGF. These views remain very diverse and experimental, which is great! I think the mix of formulas from each Badgeholder ensures a good diverse set of viewpoints.

This dilemma is whether or not we reward for retroactive impact made up to this current point, or we reward projects with potential future impact in need of funding to offset risks. Perhaps these are two different visions and may constitute two different funds with different objectives. I tried to balance both these views in how I voted, leading to an allocation of greater than 0 to a total of 183 projects, with the highest one at 1.3%.

The goods that generally made it to the top were still projects that have made a proven impact in our ecosystem including BLST by Supranational (1.3%), Keccak hashing (1.2%), Ipsilon (1.2%) and the monorepo dependencies collection (1.3%). These are undoubtedly projects that make the Ethereum/Optimism engine run and rewarding teams that make it run better, including researchers! The highest rated experimental allocation is NiceNode, getting 1.2% of my vote. This was interesting because as a Protocol contributor, I understand the great need for better node running UX in general, but NiceNode is also not a productionized project that has made a sizeable impact… yet (no pressure guys!). These projects should not be sidelined from public goods resources because they are taking great risk to solve a need in the community.

For education, the top allocations went to ETHGlobal at 1% and 0.8% each went to RadicalXChange, L2BEAT, Kernel, IC3, CryptoStats and BuidlGuidl. There were some disadvantages of my allocation methodology that made Education votes (which I’m the least in-tune with) less impactful than I think they should be, which will be discussed below.

As a Protocol contributor, I inherently have biases on what I think is useful. This experiment helped open my eyes to realize how important every layer of the stack is and the awesome projects people are building on top of what we do at Layer 1. To minimize this bias, I allocated 33.3% of my vote to Infrastructure, 33.3% to Tooling and 33.4% to Education.

Upon reviewing the results, the flaw in this methodology is that some categories are more saturated than others and also mislabelling of a category could’ve cost your project some allocation. In fact, some can even argue that a project is Infrastructure instead of Tooling, which would’ve made an allocation difference in my methodology. Due to the lack of Infrastructure projects comparatively to Tooling and Education, every infrastructure project got a greater share of allocation just because it was in a less crowded category. Perhaps a better strategy would’ve been to allocate by category ratios.

Additionally, Education is such a hard thing to evaluate in terms of quantity and quality. It also requires extensive research as to what your project/DAO/community has accomplished over the years and it’s hard to evaluate 73 of these equitably in a short period of time. I’ve recommended projects (alongside other badgeholders) to dedicate a one-pager, tweet thread or some other quick medium of explaining what they’ve developed, done or will do to educate users of our ecosystems. Many did not and those who did had a huge advantage in getting votes from me. I generally evaluated the content, quality, tools, target audience and other quantifiable metrics to gauge impact. In essence, if you had one medium of distribution (e.g. videos only on YouTube), you had less allocation than someone who put together an entire course with templates in a repository alongside their videos. Anything unique that you did scored you bonus points. I’m always interested in innovative ways to onboard new users into the ecosystem.

I made a huge effort to dig deeper in the education category because I believe it’s important and where I’m most ignorant. With tooling and infrastructure, it’s pretty straight forward what the problem is and how they’re trying to solve it with their code/product. For education, especially if courses and newsletters are mixed in with DAOs and NFT projects, makes it hard to distinguish the noise from the signal.

Part of what I believe we are solving for here is to provide resources to projects that are offering a good/service to everyone at a loss, in order to fulfill a demand in the community that isn’t expected to directly generate revenue. I think this requirement is what distinguishes a public good from a (pre)seed/angel investment. So when making allocations, I feel that:

• The most important criteria is that the good is useful or it may be useful in the future for our ecosystem. The more parts of the ecosystem it benefits, the better.

• Secondly, the usage of the good is free and will stay free for everyone. Therefore, somebody or some entity is taking a loss for running and maintaining this good.

• Thirdly, if you’re backed by VC funding or a for-profit company/product, I feel that the needs impact isn’t as great as you have access to resources at your disposal or have the ability to raise more resources.

These beliefs in how I define a truly public good may have affected the individual project allocations for RPGF.

Badgeholders are generally very active members in the Ethereum ecosystem and we have a limited amount of time to dedicate for research and due diligence (we do this voluntarily!). During the process, it was brought to my attention that there were some projects that potentially misused previous grant funds or had violated codes of conduct as the voting period was ongoing. Some badgeholders had even submitted votes before allegations were brought to our attention and I think we can all agree allegations take time to resolve. There should be some formality in how badgeholders should hastily receive these disclosures, so we can account for this in our voting. At the same time, we need to prevent the system from being abused to nefariously accuse other projects for violations with little to no merit in their allegations. I, and potentially other badgeholders, would have disqualified projects undergoing some vote, investigation or accusations with merit until matters were resolved. RetroPGF does intend to retroactively reward projects in future rounds, so missing out on a specific round until matters are resolved seems to be a prudent way to handle these governance issues.

I highly recommend that you look at other badgeholders and how they’ve allocated their votes so we can build an optimal, effective and fair way of distributing resources to public goods in need of resources. I definitely spent more time on this than I would next round, so I would need to optimize my strategy going forward. Optimism’s voting rationale post, alongside the Twitters and blog posts by other badgeholders are a great way to expand the social experiment. I hope we can figure out how to better shape RPGF3+ together and create a vibrant and sustainable public goods ecosystem for the future!