What is the cost of decentralization?

From browsing around forums and chat groups, everyone appears to want decentralization, but many don’t even seem to understand what that means. I even struggle, at times, to understand the nuances of the implications of decentralization. Are there any resources that discuss what the costs/trade-offs of decentralization are? I realize that this is a fairly general question, so perhaps I should clarify that, for now, I’m interested on more protocol level trade-offs as opposed to social aspects, such as governance in DAOs (i.e. let’s start from square one).

From a purely heuristic standpoint, I can see that there is a time cost associated with waiting to achieve consensus among a larger group. Does this push transaction fees higher? Are there other effects? Are there theories or data that show in more quantitative terms what these costs are as decentralization increases?

Just my opinion here,
The traditional costs are basically the major issues that have held back Ether and other crypto projects that Cardano is working to solve; Speed, Costs, Scalability, Interoperability, Governance, Sustainability

(I think governance will be the biggest challenge, but also the most important, and this is where I think Cardano will really stand out in the future)

Perhaps you will find something to read here: Research - IOHK

On the flip side, what is the cost centralization? (You get a fast database, but also the negative outcomes associated with concentration of wealth, and authoritarian governments)

1 Like

Agreed. In regard to my original question, I’m trying to understand why there is a push for k pools. I’m assuming there is some rationale behind choosing a certain k as opposed to just letting as many pools exist as want to exist.

I’ll probably post something soon along the lines of this topic soon. The gist is that I see a lot of potential problems with achieving a ‘good’ governance structure on a PoS network, especially before identity solutions are introduced.

I think that how you define decentralization matters. PoS means that wealth is largely synonymous with power in the network. Wealth functions as a feedback loop, so my worry is that since concentration of wealth can’t be forbidden in the protocol, it is likely to occur. If you couple this with governance then there is an even greater incentive to concentrate wealth as it translates into voting power as well.

There is a function of the protocol that rewards smaller pools, and the saturation point is a disincentive to delegate to larger pools. Therefore Cardano is designed to limit and reduce (if not solve) some of the issues around concentration of wealth. Incentivizing more pools means more decentralized, and less concentration of wealth. (in some ways, it actually can be forbidden in the protocol. or maybe forbidden is the wrong word, but by changing the incentives, you change the outcomes) There is also discussion of implementing separate voting/governance tokens, so we may see a system in which wealth or the quantity of ada you hold, is not directly proportional the number of votes you have. Yes there are many “potential problems with achieving a ‘good’ governance structure” but this is why Cardano is so focused on first principles, research, peer review, etc. It’s not a simple task, but they are employing some of the best minds in the world to research these topics extensively to formulate the best possible outcomes.

What specifically are you referring to here?

This is true, but larger pools engage in pool splitting to circumvent this. That’s why I say that how you define decentralization matters. k is a decent metric for the optimal number of pools, but it says little about how decentralized the network is. An Alternative to a0 and k

Hope this video helps to bring clarity. I don’t want to attempt to explain the entire functioning of the protocol, as I would butcher it and probably add more confusion.

There’s also this one: Ouroboros: Cardano’s Proof of stake blockchain protocol. - YouTube