Firstly, let me start out by saying plutocracies are fascist bile, and stake-weighted voting sets us back centuries. Till stake-weighted voting is consigned to the annals of history, voting on Steem will always remain abusive. Plutocracies have failed throughout history, and there's zero evidence to suggest it'll ever work for Steem. I'm open to being presented evidence to the contrary, of course. That said, the goal here is to work within that context, and give the users some options to mitigate abusive voting. This idea came off a conversation with @trafalgar in a different thread. All thanks to him, as this post would not exist without his inputs. 

Secondly, this is merely a thought experiment, and not a proposal. I don't think this is practical, but could be refined, modified and integrated into a solution that works. I know this is all over the place, and no one will read this, so if you're crazy enough to read my mad ramblings, I apologize and thank you. 

Currently, a whale can make sweeping votes. Let's consider two example scenarios. 1) The whale could make a self-vote worth $100, and 2) The whale will take a popular post down to 0 by downvoting it due to some personal vendetta. 

There could be 1,000 minnows each with a $0.1 vote who could downvote the self-vote and upvote the downvoted post, but currently, no one will do so because of a) Fear of retaliation, b) they don't want to waste 2% VP, and c) no one else will do it, so it'll be a waste of time. 

The idea suggested here is the ability to counter a specific vote. Dan Larimer had first suggested vote countering back in 2016 when he was the CTO of Steemit Inc, but this idea will be more granular than a simple 1:1 vote counter. 

Each vote will have its own R-shares distribution, just like there is for a post today. People will be able to counter a specific vote, while others can choose to support the vote. 

In scenario 1), the counter vote will be a downvote; while in scenario 2) the counter vote would be an upvote. Please note that by countering a vote on a post, the voter forgoes the ability to vote on the parent post itself. I.e. in scenario 1), you're countering the whale's self-vote, but you will not be voting on the whale's post where they left the self-vote. 

Speaking of curation rewards, we'll ignore that for this post for the sake of brevity and clarity, and instead talk about a different incentive - VP costs and subsidies. 

Like how curation rewards work today, there's a Wgt% assigned by how early a voter is, and the R-shares allocated by their vote. The exact formula of Wgt% is TBD, of course it's not going to be like curation rewards. The system will need to be modified to account for -ve R-shares (i.e. downvotes) as well as +ve R-shares (upvotes). 

The end result would be a distribution algorithm where voters are some magnitude of "correct" and "incorrect", and some magnitude of early and late. Depending on a) how correct (or incorrect) they are, and b) how early they are, voters get a VP subsidy (or cost, if incorrect). 

The median VP cost will always remain 2% VP (or relative proportion of, considering voting strength %). At the time of voting, each vote will be debited 2% VP, as it is currently. The subsidy or cost over/under 2% will be incurred when the parent post pays out. Till then, all voting is considered pending, as it is currently. 

Let's consider scenario 1), regarding the whale's $100 vote. To be clear, I'm not suggesting that there's anything wrong with self-voting. But maybe 2,000 minnows each allocating $0.1 feel otherwise. 

Let's begin with 1,000 minnows deciding to counter the initial $100 vote down to zero. In this case, the 1,000 minnows and the initial whale are both half correct and half wrong. However, by being the earliest voter, it is more like the whale was half wrong by making a contentious vote in the first place. To that end, the whale will have to pay a penalty in VP over and above the 2%. Meanwhile, the 1,000 minnows will get VP bonus to partially subsidize their 2% VP cost. How much the bonus is will depend on how early they were to predict the "correct" outcome. In this scenario, the first few minnows may even stand to gain a VP bonus of above 2%, effectively having earned VP from the vote, instead of spending 2%. However, those piling on the downvotes once the outcome is more certain will likely end up with a much smaller subsidy, and end up incurring a cost near 2%, but marginally under. 

What if another 1,000 minnows join in to downvote? Now, the post will stand at cumulative -ve R-shares equivalent of -$100, signalling this was an extremely bad vote. When this happens, the whale will start incurring increasing penalties, while the earliest minnow counter voters will start seeing increasing VP bonuses. 

Of course, there will be those that'll support the self-voting whale, and will vote with + R-shares on the vote.

Similarly, for scenario 2), people would be countering the downvote with upvotes. It doesn't matter if the vote being countered is a downvote or upvote; all principles apply. The potential difference is people are unlikely to downvote beyond $0, while upvoting past the $100 downvote is more likely. This will need to be addressed in the distribution algorithm. 

At the end of the payout period, we'll end up with **a distribution of each voter being some magnitude of "correctness" and timeliness**, include the initial vote that's being countered. Some voters will pay a penalty, while some will receive a bonus. The best curators may even stand to gain a VP >100%, effectively incentivizing good quality control. 

Of course, this needs to be combined with further incentives and a cultural shift to normalize countervoting. As it stands now, everything I have said above is not going to work.