As the third version of the renowned RetroPGF public goods funding program, RetroPGF Round 3 (R3 for short) announced its results not long ago. 501 builders, writers, creators, educators, and contributors from across the Optimism ecosystem (out of 643 candidates) received 30M OP as rewards for their contributions and impact within the Optimism ecosystem. The following insights are extracted based on an exploratory analysis of R3’s public data.

An Overview

The 643 candidates in R3 mark a whopping 230% plus increase over R2’s 195, whereas the number of badgeholders only increased from 70 to 146. There had been voices about how the voting process was ‘overwhelming’ during R2, how will R3 fare with such an asymmetrical increase?

Not bad, actually.

With the introduction of ‘list’, badgeholders can assemble their own candidate lists like they would their playlists, as long as there’s a reason. It can be ‘they’re all AA wallet projects’, ‘they create Spanish content’, or even ‘their logos are red’ (I made this up). This way, badgeholders can examine and reward multiple candidates in one go.

Like previous rounds, R3 took the median of all OP allocations for each funded project to calculate its weight in the total funding. All candidates must declare their contributions and impact within the given period as detailed and quantified as possible. All badgeholders were then given one month to review all candidates before casting their votes and allocating a certain number of OP to several candidates. Their goal is to evaluate a project’s impact and try to make up the difference between their actual profit and the profit they deserve.

Preliminary Analysis

There are 4 non-exclusive categories in R3: OP Stack (146 funded out of 165 candidates), Collective Governance (104/104), Developer Ecosystem (252/304), and End User Experience & Adoption (355/472).

A total of 531 candidates submitted their applications as ‘projects’, of which 429 received funding, funding rate (80.79%not to be confused with ‘funding rate’) is significantly higher that those ‘individuals’ (72/112 = 64.29%).

A visualized distribution of ballots and its correlation with OP received is shown below:

All candidates (however unlikely they may look) received at least 4 ballots.

Apart from several outliers, the scatter plot shows a somewhat consistent correlation between ballots and number of OP received, albeit not consistent enough to tell us that all ballots are created equal.

Because they’re not, even if we pick only the median ballots. The rule of thumb for R2 is that the number of OP received equals 1,000 to 2,000 times the number of ballots received. But there are still some exceptions that goes as high as over 7,000 OP per ballot received.

Of all categories, OP Stack, Developer Ecosystem, and End User Experience & Adoption all boast similar OP per ballot received, while Collective Governance candidates get significantly fewer OP, even though all candidates within this category received funding.

Did They All Say That?

All candidates were required to submit a summary of their contributions and impact within the Optimism ecosystem. After a preliminary analysis of these paragraphs, the 25 most frequent words in all candidates’ contribution statements are (removing words like ‘the’, ‘is’, and ignoring ‘Optimism’ and ‘OP’):