The Economic Changes in Hardfork 0.3.0 and Their Rationale

The Economic Changes in Hardfork 0.3.0 and Their Rationale

I would like to make some further comments regarding the recent suggestions I made on the Blurt economic system, and the decisions taken as to what will be implemented in the forthcoming hardfork.

I feel it is important that members understand the decisions taken and the path going forward. I shall also continue to post about cryptoeconomics in general, including research findings and systems implemented on other chains. Deep down, designing an economic system is not easy, indeed it remains an open and hence unsolved issue. We are thus not just developers, but also explorers and experimenters.

Many cryptocurrencies do not have an internal economy such as Blurt's; most have the much simpler case of miners and mining pools, and some are developing financial systems at a level above the chain itself. Very few have an economic system built into the blockchain itself. Understanding the logical and mathematical issues involved should help members appreciate that this remains a work in progress. One thing that worried me deeply about the various changes made on Steem was the lack of truthful analysis and the seeming lack of care about consequences - those second and third order effects that come and bite you from behind! I shall write about these soon, but let me not get diverted too much now.

OK, let's get back to Blurt! You can see on Gitlab the decisions taken regarding the reward curve and the vote dust threshold.

The dust threshold was uncontentious and will mean that all users shall see their votes worth a bit more. This will be especially welcomed by those who are currently seeing their votes worth 0.000 BLURT. Also note that this constant was never removed in HF21, so that the estimates of the so-called reward curves on Steem and Hive are all slightly wrong.

The reward curve needed some more thinking. You can read my analysis here and I appreciate the inputs from various people, whether as comments or within chatrooms. What struck me was an issue that I hadn't really factored into my original analysis: the increase in vote farming. Votes will already see an increase due to the removal of the dust threshold, but adding another significant increase before solving the issues around bad actors would make such issues even more worrying.

Hence, the decision was taken to just change the so-called CONTENT_CONSTANT within the hyperbolic-linear formula, making it lower and in line with the recalibration of 1 BLURT equalling 1 Vest. What this does is make the larger votes attenuate much faster to true proportionality. What users should experience is that those larger votes, such as from community curation accounts, will be slightly larger. This means that those members with modest accounts and good quality posts will see an increase in earnings even if their own votes show only a modest rise. Remember that the reward curve applies to each post as a whole, so this change aims to reward quality posts while keeping low-level vote farming as a minor economic irritant.

Once there is a robust system to deal with bad actors and voting abuses, then we shall revisit the reward curve and make any changes then needed. I would also like to analyse how the Blurt reward pool reacts to HF 0.3.0, so it is also prudent to change economic parameters bit by bit so that we can have more predictive confidence in further changes. I would also like to analyse the curation rewards curve as that has been creating some unexpected rewards - I shall talk about this at more length in another post.

Blurt's HF 0.3.0 has many more parts to it than the economic matters I have been entrusted to look at, and I hope members will see the whole package as moving in the right direction.