Hello. We conducted a survey on Bitcoin’s 2nd layer and submitted the attached paper to a workshop three months ago. However, after submission, we came to believe there were several errors, and we officially withdrew the paper before the acceptance decision. This was because we thought the following parts were problematic:
-
The definition of decentralization was oversimplified
-
Bitcoin’s 1st layer mining is also biased toward corporate characteristics
-
Even if individual Ethereum Rollups are centralized, there’s a possibility they could be called decentralized if they can be interconnected
Over the past three months, we’ve been thinking about things like:
- Whether decentralization is meaningful in terms of the diversity of attributes of people running nodes
However, we haven’t been able to understand the decentralization clearly and couldn’t rewrite it.
We’re posting this here hoping to receive feedback. We would appreciate feedback you might have.
https://github.com/kurihei/delving/blob/83a6a0c57aca21febe0e0d060fbe8852094a0871/delving.pdf
It may be helpful to think of decentralization as a measure of resistance to attack, rather than a unit of measurement itself. Centralized systems tend to have more opportunities to exploit an attack, whether by governance or physical action. Decentralized systems do not share such vulnerabilities.
In the case of Bitcoin, the network is difficult to attack due to its decentralized properties. Of course, we cannot measure attacks that are not happening/not successful since you cannot prove a negative. However, an attack on a centralized point of the network is easy to recognize (Example: Marathon Digital Holdings Becomes the First North American Enterprise Miner to Produce Fully AML and OFAC Compliant Bitcoin : MARA (MARA) )
Hope this helps.
The best definition is from Hurwicz:
A process is defined as informationally decentralized when each economic unit (such as a consumer or firm) initially only has information about itself (its preferences, technology, or resources) but not about other units. The process then requires an exchange of signals (called messages) to attain objectives such as efficiency of the system as a whole.
To this must be added the lack of a “credible enforcement anchor” – the lack of a central player who can coerce or compel participants to follow strategies they may not follow in the absence of credible enforcement.