I shared a few additional thoughts in the citizen’s channel on the voting algorithm that was used in this round (and in the previous round as well), and I’m copy-pasting them here for future reference:
Would it make sense to not take the median of the percentages allocated to the projects in a category, but rather the median for each project of the absolute OP values each voter has assigned, based on the category budget they have suggested?
Ie.: Let’s say we have a round with categories A, B and C. I am asked to vote in category A, which has projects a, b, c and d.
Now, let’s say we have all initially settled on an overall round budget of 3M OP.
I might then suggest allocating 1.2M OP to category A, and distribute them as follows: 25% for a, 30% for b, 5% for c, and 40% for d.
Converted into absolute values, that would be 0.3M OP for a, 0.36M OP for b, 0.06M OP for c, and 0.48M OP for d.
Now, take the OP allocations of all voters for every project in the round and compute the median for each project. Normalize to fit the overall agreed-upon round budget.
Would that work?
I don’t know if I’m missing something important, but it feels like that would make it much more likely that the result is actually reflective of the voters’ wishes.
As it is now (RF5 and RF6), I may have an idea as to how much each project in my category deserves, and I might budget and distribute accordingly (using percentages), but my percentages may end up yielding vastly different results than I had anticipated if other voters decide upon a much larger or smaller round budget, or if my category is getting a much larger or smaller percentage than I had expected.
For example, in this round there are a few very small projects in my category. They are good projects, though, and I would be happy to give them 1-2.000 OP as an incentive for others to take such initiatives. However, if my category ends up getting only half of what I expected, my 1500OP allocation turns into 750OP, and the applicant gets nothing (because of the threshold). And if my category gets 3 times the amount I expected, the project gets 4.500 OP, which would be too much and maybe more than a more deserving project in a category that ended up getting a smaller cut than another voter of that category had anticipated…
As a result of this design, I am currently doing A LOT of “what-if” thinking… Like, how will my vote work out if other voters vote for a much larger or smaller overall budget? What will happen if they vote for a larger budget AND vote for my category to also get a larger share? What if…?
I think the alternative design I have suggested above might be more simple and lead to more conviction voting.