RE: The horrendous failure of curating Quality Content

Either way, I'd love to discuss possible solutions, do some simulations through them to see weaknesses and how they could potentially be abused as well

I appreciate that.

Without question, any proposed changes need to be thoroughly examined for potential weaknesses and vulnerabilities to various attack vectors prior to implementation.

---

The more I've thought about it, the more I am drawn to solutions that ultimately empower the witnesses. The witnesses are the ones we are predominantly relying upon, to safeguard against hostile forces.

My basic thought here is that giving a strong lever to the witnesses (e.g. enabling them to quickly ramp up or down the relative strength of DVs) would preemptively discourage potential bad actors from even attempting to game the system on a large scale, because their efforts could be squashed in very short order, i.e. by a majority (or super-majority) of the witnesses adjusting the controlling parameters.

My current suggestion in that regard would be for two new witness parameters to be added via a future HF: dv_strength and counter_dv_strength, with a max and min for both hardcoded into the HF.

Every time rewards are distributed from the rewards pool, if there are DVs or Counter-DVs associated with a given post, then those values get modified by the median values from the current top-20 witnesses for dv_strength and counter_dv_strength.

Whenever dv_strength == 1 and counter_dv_strength == 0, that would represent today's status quo.

This would allow some experimentation with very limited risk (because the witnesses could quickly restore everything to the current status quo if any problems were to arise).

For example, the witnesses could start by slowly reducing dv_strength from 1.0 to progressively lower levels.

As long as community-wide anti-abuse efforts (related to plagiarism, spam, etc.) remain strong, then the potential for DV abuse keeps being reduced, without an increase in other forms of abuse.

However, if and when other forms of abuse begin to ramp-up, the witnesses could respond by quickly ramping up dv_strength.

---

Just implementing dv_strength as a witness parameter would be a significant improvement, imho.

Even so, I would like to see free "Counter DVs" enabled as well. It would be harder to implement (codewise), but it would help democratize DVs and make them much less toxic. People who receive DVs without any significant Counter-DVs will know that those DVs were 'deserved'. People who receive DVs then get loads of support from Counter-DVs can feel vindicated. If other people wish to pile on more DVs to counter the Counter-DVs, then so be it. As long as it's a community affair, rather than simply 1 or 2 whales acting unilaterally, then that's all worthwhile engagement, imho.

In fact, there might be some front-ends that would tweak their trending algorithms to specifically elevate posts that are receiving a lot of DVs and Counter-DVs, because it would promote posts that lots of individuals are passionate about (on both sides of a given issue).

For a Counter-DV system to work, the strength of Counter-DVs would need to be significantly less than the strength of DVs. I am thinking 25% to 50% might be ideal. However, as a witness parameter, the witnesses would collectively decide what's best.