# RPGF3 Considerations *Median Votes, Transparency, and Future Improvements* By [Routescan](https://paragraph.com/@routescan) · 2024-06-26 --- ![](https://storage.googleapis.com/papyrus_images/06d594ba31e7740f2220c1c513093e57.png) Introduction ------------ The Optimism Collective’s third round of **Retroactive Public Goods Funding (RetroPGF 3)** occurred in the fall 2023, allocating **30 million OP tokens to reward** contributions that have supported Optimism's development and adoption. For detailed voting and badge distribution results, visit [Optimism Governance](https://gov.optimism.io/t/retropgf-3-voting-badge-distribution-results/7001). In RPGF3, 643 projects were evaluated by 146 judges, who voted to determine the amount of OP tokens to be granted for previous work. The voting results were determined using the median of all votes for each project. 1\. Median and Zero Votes ------------------------- In the crypto community, there was a significant discussion about the impact of zero votes on the median voting results. Not everyone fully understood that using a median instead of an arithmetic mean zero votes would affect the outcome. Analyzing the top 20 projects, we found that up to 14% of votes were zeros, gradually decreasing to zero. ![](https://storage.googleapis.com/papyrus_images/093705e85e34f05547ace46502b3877e.png) Distributions of 0% of votes per project. Another consideration is that the projects related to Bankless were probably the ones that received the most zero votes; this becomes evident if we group them all together. ![](https://storage.googleapis.com/papyrus_images/0a9b44c604021dbb059aa154071b072d.png) Distributions of 0% of votes per project and with all Bankless projects merged into a single one. This is likely because the judges did not perceive that each individual Bankless project participated independently. Here we report the specific thread on the topic of [Bankless projects](https://gov.optimism.io/t/clarifying-bankless-academys-individual-application/7160/12). Since the voting was anonymous, we cannot identify who cast zero votes or their reasons. However, the repeating percentages across multiple projects suggest that a specific group of judges may have cast zero votes on multiple projects, either due to misunderstanding the median use or as a deliberate strategy to penalize certain projects. This hypothesis can only be confirmed by the Optimism Foundation, which knows the votes and voters. If true, we hope for adjustments for future RPGF rounds to prevent excessive penalization due to simple misunderstandings. For more details on the voting and badge distribution results, visit [Optimism Governance](https://gov.optimism.io/t/retropgf-3-voting-badge-distribution-results/7001). Below is a graph showing the difference between the median and the median calculated without zero votes. We can observe that the most penalized project was Synthetix, where out of 54 votes, 5 were zeros, bringing the median from 150k without the zero votes to a median of 100k. Next, we find Solidity, where the difference is more contained because it received the most votes overall. ![](https://storage.googleapis.com/papyrus_images/1b6af3940b5a09a0615e4d6a6c6e41ad.png) Difference between median and median without considering 0 votes, negative values means the median without zeros votes is higher than the median used to distribute RPGF3 grants. 2\. Anonymous Voting -------------------- As mentioned earlier, the voting process in **RPGF3 was anonymous**, however, some individuals made their votes public, for example **Jesse Pollak**. Pollak, the creator of Base, disclosed his intended distribution of OP tokens before the voting began. We aimed to determine whether Pollak's transparency influenced other judges. Firstly, it is important to note that an approximation was done. The data shared by OP showed votes as whole numbers, so Pollak's votes were approximated as whole numbers too. ![](https://storage.googleapis.com/papyrus_images/14314e16629b403d25a6de366287da1b.png) Difference between median without considering Pollak's votes and median, negative values means the median without Pollak's votes is higher than the median used to distribute RPGF3 grants. Pollak voted on 69 out of over 600 Dapps. The image above shows the difference between the median excluding and including Pollak’s votes. The Mirror project is the most penalized by the votes given by Pollak, while on the contrary, Solidity is the one that benefited the most. Furthermore, it is possible to say that: * For 43 out of the 69 Dapps, about 60%, the median vote was exactly the same as Pollak's. * For the remaining 40%, the average deviation was approximately 500 OP tokens with a standard deviation of around 6000, which is roughly a 2% difference from the actual vote. When comparing the average votes with and without Pollak's contributions, the average increased slightly by about 1000 OP tokens with a standard deviation of 2000. ![](https://storage.googleapis.com/papyrus_images/b3605c081cad2cc807a189c2b2962139.png) Difference between mean without considering Pollak's vote and mean. This indicates a correlation between Pollak’s votes and those of other judges, suggesting that Pollak's disclosed votes might have influenced some of the other judges' decisions. For a detailed look at Pollak’s voting data, you can refer to his [spreadsheet](https://docs.google.com/spreadsheets/d/13_omIPXbvEukklwIVK2lm-lbf1W3BfQEOfRxWp6ij5k/edit#gid=638856714). Conclusion ---------- In this analysis, Routescan aims to shed light on some of the complexities and potential issues that arose during the third round of Retroactive Public Goods Funding (RPGF3) by the Optimism Collective. The discussions around the use of the median for voting and the impact of zero votes highlight the need for a clearer understanding of the voting process among participants. The presence of zero votes, whether due to misunderstanding or deliberate strategy, has significantly influenced the outcomes. This suggests that for future RPGFs rounds, it would be beneficial for the Optimism Foundation to consider adjustments to the voting mechanism to ensure fairer results and prevent projects from being unduly penalized. Furthermore, the analysis of Jesse Pollak’s publicly disclosed votes and their potential influence on other judges underscores the importance of transparency and its effects in the decision-making process. While Pollak’s votes were largely in line with other judges, the correlation suggests that transparency can shape the overall voting behavior. Routescan believes in the value of a free market and transparent processes. By addressing these issues, the Optimism Collective can enhance the integrity and effectiveness of its funding mechanisms, ultimately fostering a more robust and innovative DeFi ecosystem. We look forward to continued improvements and contributions from the community to refine these processes. For more information on the RPGF3 voting and badge distribution results, visit [Optimism Governance](https://gov.optimism.io/t/retropgf-3-voting-badge-distribution-results/7001). **References:** * [Vote Summary](https://vote.optimism.io/retropgf/3/summary) * [Delegates](https://vote.optimism.io/delegates) * [Jesse Pollak’s Voting Data](https://docs.google.com/spreadsheets/d/13_omIPXbvEukklwIVK2lm-lbf1W3BfQEOfRxWp6ij5k/edit#gid=638856714) * [Bankless Projects](https://gov.optimism.io/t/clarifying-bankless-academys-individual-application/7160/12) --- *Originally published on [Routescan](https://paragraph.com/@routescan/rpgf3-considerations)*