Discussions on pricing model
- Sergio from EOS Argentina:
Proposed a different approach to the pricing function, assimilating with what is currently used in automated market makers. The intent of this proposal is to remove maximum price cap and assure greater variability in price, especially within the higher end of utilization(
u=[0.9, 1.0]
), without imposing exponent
parameter too high.
- Aarin from B1 (Originator of PR #397):
Although the proposed function would do its intended job, he's not sure whether whether the practical benefits justify the development time of implementing a very different model (researching about trade-offs, etc.), while existing model of his would just do its job. Thus, at the moment, he thinks it is safe to say that getting sufficient variation in price at the high end with a higher
exponent
in the existing model and accept the limitations at the lower end would be a good way to begin with. (Details some of the theory and math can be found here: 1, 2, 3)
- Yves from EOSNation:
Different parameter settings can be tested on different testnets, noting that it would be a safe assumption that demand will remain high regardless due to mining.
- Bo from Atticlabs:
Proposed an alternative model that calculates and checks at the beginning of the
rentbw
method that the maximum cost exceeds the minimum cost.
- Exception from GenerEOS official feedback on 25 Sep, 2020.
The best ‘price curve / price curve family’ discussion:
EOS Rio in [5] suggests that a change in the pricing curve would further reduce wasted resources by keeping resources more affordable at the higher utilisation end of the curve.
While we agree with EOS Rio that there might be some optimisations required for the price curve, we believe that the price curve proposed by block.one is suitable to achieve the goal of keeping resources affordable up to a resource utilisation of ~90%. Having said that, there is not enough data available to give definitive predictions on resources pricing and more research needs to be done in this area.