You are viewing a single comment's thread from:

RE: HF21 (and 22) are done, how do you feel?

in #hardfork5 years ago

Looks like we have a similar understanding, but different priorities. That's fine as long as we get there before too long.

But I would add that I'd like to see numbers. I've been asking for numbers since the announcement of EIP in HF21 and haven't seen them. It's too late for those to matter now, I guess (though I'd still like them... what percentage of the rewards pool was going to abuse?) The new numbers I'd like to see are:

Are middle earners rank and file? Which is to say, do most top level posts earn $4-10? Or is that actually still a part of the top echelon? Is the freewrite community an outlier, or are we the norm? I think we're the norm, but it's hard to know, but if we are, there must be a lot of regular users out there who are getting $0.20/post. To me, if that's more than 50% of the crowd, that's the rank and file.

The curve exists to serve a purpose. Punishing responsible low-earning users isn't the intent at all, but is an unavoidable side effect.

I know it's not the intent, and I know the purpose, but is the cure worse than the disease? ATM, I'd say yes, especially considering that we didn't try the EIP without the curve first, which is what I'd like to see. Boy howdy.

I'm glad the downvotes help. They'd help even more in a linear reward curve. But sure, more curation. I'd even love to see a whale make a whitelist of "responsible low-earning users" to give those $0.20 upvotes that get them back to where they were. That'd go a long way to ameliorating the unavoidable side effect.

Sort:  

We didn't try EIP without the curve because our best available analysis and the consensus view of stakeholders, witnesses, and the developers is that it wouldn't work. (This does not mean 100% that it wouldn't work. but we have only a certain number and frequency of shots at this so we have to just use our best judgment to choose shots and take them.)

The limited amount of free voting power and the fact that there isn't any known way to compensate downvoters means that relying too much on downvoters doing a lot of hard work is a very dodgy proposition. The more obvious, more concentrated milking and abuse in Trending and voting circles is one thing. Digging through tens of thousands (and potentially more than that if incentives change) of tiny milking payouts is something else. The curve takes the place of someone needing to comb through and enormous number of low-payout posts and individually decide which are milking and which are not. That it is a broad brush is the unintended side effect, but, unfortunately, we simply don't know of a better way.

As far as giving low-earning voters blanket free upvotes simply for existing, I'm not in generally favor of that. It falls within the scope of curation to decide what rewards are appropriate and to a large extent that is independent of whatever rewards might have been previously (under a very broken system that, broadly speaking, was widely viewed by stakeholders, witnesses, and developers as a failure that was contributing to Steem's decline and not its success, though this does not mean that every single payout under the previous system was bad; that certainly was not the case).

As I said earlier, increased curation is still ramping up and will likely take some time. Most probably some low earning users will be picked up by increased curation (including increased votes that may be given because rewards are considered deserved by the curators and the curve requires it) and some will not, which is how this is supposed to work.

Well, we continue to disagree. I'm aware of the reason for the curve, but I still haven't seen any sort of numbers to evaluate the theory by. I don't know how much of the reward pool was taken by the behaviors CLRC is designed to solve.

That it is a broad brush is the unintended side effect, but, unfortunately, we simply don't know of a better way.

I know it's anathema to this crowd, but KYC would solve this problem. One person, one account would restrict the number of accounts bad actors could have. You could even have a two-tiered system, with opt-in to KYC that gave users access to the linear rewards curve, and opt-out that only gave you access to a curve that had the drawbacks of CLRC without the benefits above that (but the same payouts as linear gave).

Obviously, setting up KYC is a complication in itself, how would the blockchain handle that status switch? Who would be responsible for verifying documents, etc. etc. But it could be done, and if done, it'd be a better way.

I don't know how much of the reward pool was taken by the behaviors CLRC is designed to solve.

Retrospectively is not the point. The point is to build a system that has the greatest likelihood of working going forward. More specifically there was very little need to milk via small accounts previously because milking using the very biggest and accounts, or whatever size happened to be convenient, worked just fine with no downvotes in use practically (<0.1%)!

but KYC would solve this problem

With enormous costs and requiring an infrastructure that doesn't exist. People already complain about 0.50 worth of Steem paid as a transaction fee to create accounts, which is tiny compared to what any sort of robust KYC costs so, no, I don't think that is feasible.

KYC is not only a matter of being anathema to a crowd but also economics. There is a reason companies like Facebook with nearly unlimited resources (at least compared to Steem!) still have a very high percentage of unverified and fake accounts, and they did even less verification (basically none) when they were trying to grow. It is just too expensive and too big a barrier to growth to do strong verification.

I wouldn't entirely rule it out going forward when the economics make sense but it definitely isn't a solution for Steem at this point.