I strongly agree with @joanbp and @kaereste points above.
I support the Dev Tooling RF Mission Budget, as we have promised it for almost a year now, it’s important we follow through for those builders.
I do not support the Onchain Builders Budget.
We gave them the most out of any group since RF3, just a few months ago… it seems weird to do a 10M round, just because it was the best rated experiment. IMO the reason it was the best rated experiment is because the distribution was the least flat. The outcome of the projects ranking actually was on par compared to RF5 and RF6 IMO (Who thinks Layer 3 should get 500k OP in RF4 really?). The big difference was that the voters didn’t have to give an OP amount, and the distribution followed the impact metrics distribution.
Does that mean we should only use metrics… NO.
I have been saying this since RF3. We can still get the benefits of qualitative assessment without the flat distribution, if we set the distribution in advance and make it competitive.
1st place gets x
2nd place gets y
etc
And (warning spicy take) IMO the bottom 50% should get nothing!
This is how most contests are done AND Impact seems to generally have a power law distribution, so the rewards should as well…
The biggest problem with RF6 was that the applicants knew that there would be a flat distribution so the projects that strategically made multiple entries took home way more OP than the ones that made just one entry e.g. Pairwise should have made a proposal for RF3, RF4 & RF5, I am confident we would have gotten 2x more OP than making 1 proposal. SIGH.
If the distribution was set to a power law distribution from the onset, the projects wouldn’t have been strategically advantaged to split up their proposals, because it’s in their best interest to WIN, to make the MOST impact.
That’s the incentive alignment we need.
It feels like we are trying to reward everyone for participating… its like 2nd grade soccer in the US where everyone gets a trophy. It’s not rocket science. That doesn’t incentivize the best to be the best. We need to change that and then we can still incorporate qualitative analysis.
I would love to see what the results of RF5 and RF6 would have been if we took the relative distribution from RF4 and super imposed it on RF5 and RF6 and how the citizens would have felt about those results.
Anyway end of rant.