You're describing is how it is now. What I described above is a way to somewhat disincentivise bad voting by adding a VP cost to it. Also, you have quoted a different solution, and not what the above is about.

Do you really think people are going to listen to some rant?

Thanks for reading. This is just a thought experiment, I wouldn't consider it a proposal. I'm glad that it piqued the interest of someone sincerely looking to improve Steem, and that's all I was hoping to achieve. 

Like I mentioned in the other thread, I'm in support of the EIP, conditional on the exact details and numbers. I believe it'll only bring marginal gains until the fundament flaw is addressed (i.e. plutocracy), but it's the best short-term proposal we have thus far and marginal gains are better than zero gains. 

Since you have my attention, I'm going to write up a few examples shortly.

While thinking about this further, I may have thought of a more specific solution to Steem's problems. The first clarification should be that curation rewards will continue to work as it does now, with the possibility of adding curation rewards to downvotes. For now, I'm going to write about simpler solutions and use existing systems available to us. 

Idea #1: **In short, the ability to counter specific votes.**

Currently, each vote assigns a certain number of R-shares, +ve and -ve. +ve votes come with a curation rewards share, depending on when they vote as per the squared curation rewards curve and their R-shares. Let's call it Wgt%. So, for each vote, we have R-shares and Wgt%. Now, in addition of a per-post basis, be a similar curation Wgt% for each vote for each countered, or contended, vote. 

Example: A whale self-votes with 1,000 R-shares. 1,000 minnows have 1 R-shares each, but today they are not going to do anything about it because a) fear of retaliation, b) don't want to waste 2% VP (proportional to voting %) for zero effect and b) no incentives to do so. 

A distribution algorithm needs to be worked out, of course. The whale will start to lose VP depending on how heavily the vote is countered. Meanwhile, minnows will be subsisided depending on how much they counter it. The ones who counter early will see more of a subsidy as per the Wgt%. 

In a perfect scenario, the 1,000 minnows have countered the vote, and the net vote now stands at 0 R-shares. The total equilibrium VP of the vote can also stand at 0. So, the whale will subsidize all minnows' voting in VP. How much they lose will depend on the exact distribution curve used. Of course, the minnows who countered early may even stand to gain VP, while those who piled on the counter votes late will probably lose some VP, but a bit less than the 2% they do now. Once again, these are just examples, a fair distribution curve needs to be worked out. 

A similar situation would be a popular post with 1,000 R-shares, which one vindictive whale reduces to 0 with one fell swoop. People can now be incentivised to counter this abusive downvote. 

These are extreme situations for illustrative purposes. A more realistic scenario would fall somewhere in the middle. People can support or counter a contentious vote. If the vote is only e.g. 25% countered, the median still falls within the original vote's side, so those voting that way are on the "correct" side. Even so, it was a contentious vote, and they'll have to pay a smaller penalty in VP. The 25% that countered do have somewhat of a case, so their VP costs will be slightly discounted under 2%. Again, depending on the distribution curve, there'll be a certain threshold beyond which the whale's vote becomes increasingly contentious and as will their VP penalty. The counter votes' VP costs will increasingly approach zero, and for some who were the earliest to counter, it will be a gain of VP. Thus, incentivizing people to get the ball rolling on countering an abusive vote. 

Important clarification: If one counters on a given vote on a post, they will not be able to vote on the post itself. 

----

Idea #2: Now, back to a more general system, as per my initial thought experiment. 

For each post, it's a competition between +ve R-shares and -ve R-shares, or at least should be. Now consider the Wgt%, which is already a system built-in to calculate curation rewards. We need to modify it so it also considers -ve R-shares. I'm aware that's more complex than it sounds, but there are econometrists and developers far smarter than me who can figure it out, I'm sure. A radical approach would be to have negative curation rewards, in this case, so bad actors stand to lose SP for truly bad curation. But we're not doing that, and restricting losses only to VP here. 

Another way to do this would be to not consider R-shares for Wgt%, just purely when they voted, empowering minnows even further. By including R-shares, it'll encourage whales to get involved with correct curation, but on the other hand, whales will continue to dictate what is "correct". I don't have enough information to know which way is better. 

Either way, the end result is we end up with a scale of -ve Wgt% to +ve Wgt%. Of course, the algorithm and formulae for calculating this distribution needs to be worked out. It'll probably be some form of normal distribution that'll be designed to identify outliers. 

Unlike the counter voting method, the median voting cost might need to be above 0 as the post is still allocating net R-shares, maybe 2% as it is now. I'd prefer a floating median, based on how divided a certain post is. If the +ve and -ve sides are equally divided, I'd imagine the median VP would need to be 0. I don't really have a good sense about this, so let's move on. 

I'd expect most posts will end up with an expected distribution, with most people ending up spending approximately the same as median 2%. It'll only be some (if any) outlier votes, that'll be penalized with VP costs, their penalty will be reallocated to the "correct" curators proportionately by Wgt% whilst keeping the weighted equilibrium at 2% overall. In a nutshell, wisdom-of-the-crowds, penalize the outliers. Of course, the crowds could very well be wrong, but we have to make that assumption here, given we are allocating a common reward pool. Furthermore, the incentives are now there for people to appeal for more votes to correct an injustice. The big problem, as I alluded to earlier, is that the stake-weighted function will still remain, and at some point the distribution/algorithm will break and a collusion of whales can end up on the "correct" side. There's a more complicated solution to mitigate that, involving maintaining relationships between voters, but that's for a different discussion. 

Example: Same scenario as above. Instead of countering the specific vote, the 1,000 minnows will now downvote (or upvote) the post. A distribution will form, based on the nature of votes. In this case, there'll be a clear outlier, which will need to subsidize the VP costs on the "correct" side. Those with Wgt% end up with greater subsidies. The tippy-top curators will end up with a net gain in VP. 

Of course, usually it's going to be some combination of upvotes and downvotes. In most cases, it's probably going to be pretty even, with most curators spending between +/- 0.2% of the median 2%. It's only the contentious ones where we'll see some outliers begin to form.  

----

Thinking out loud, a "quality control" curation project can be formed. With the right coordination, we now have a group of people that are incentivised to work together and start downvoting bad bid botted (tongue twister!) posts en masse. With enough of the right decisions, they could end up with 100% VP (or even greater than?), while there'll be long term damage with even whales perpetually ending up with cripplingly low VP. The whales will have to either curate better or exit the platform. In the 3 years I have been on Steem, nothing has been as exciting as the formation of Curie. A quality control curation project will get me excited about the platform again. 

Of course, all of this will need a cultural shift. Downvote subsidies as per the EIP, other incentives, other ideas to get the ball rolling. Ultimately, the best solution will need to take the best of all available ideas. 

Anyway, this has been all over the place, sorry. I hope this helps in any way. Even if not, thanks for reading!