Author: Adam Back 2013-11-07 20:05:34
Published on: 2013-11-07T20:05:34+00:00
The paper discusses the limitations of a model that does not take into account the fact that when another pool mines a block, the selfish pool will not be able to change the existing winner's mind. The miners already have an incentive to maintain fast and secure links to other significant miners. Additionally, broadcast is not instantaneous, which tends to erode the achievable gamma parameter. The profitable threshold alpha is estimated to be in the range of 25%-33%, assuming whatever techniques to reduce latency are used by the selfish pools can be used by other pools.The main result of the analysis is that even with gamma=0, it is still possible to win once the selfish pool reaches 33%. However, it is not clear what will happen if multiple selfish miners compete with each other, as there is a risk of mutual sabotage. Despite this, it seems that reducing the use of and motivation for pools could be a constructive direction for reducing centralization risks. The source code for the simulator may be useful to release, and some have suggested that Peter Todd observed something similar regarding miner incentives some months ago. While analysis is welcome, it is important to consider other profit motives in bitcoin that are not exercised, such as the restraint exercised by miners who avoided growing over 50%.
Updated on: 2023-06-07T19:57:57.618758+00:00