Of course, people will always try to maximize the system. But the entire point of this system is to make sure, that the top stake holders don't have as much influence on the top 20 witnesses.
Especially since witnesses 1-20 have the power to make all the decisions in relation to hard forks etc.
With >= 20 votes, pumpkin has an enormous influence in this. If we reduce this to <= 10, this helps a lot.
I'm not talking about saving the world, but I'm talking about decentralizing the election. And reducing the number of votes, certainly does that.
And I agree, this can be further reduced, to 5,3,1. But I think 10 is a good number. 10 allows people to select a number of witnesses which represent them well and they think represent Steem well while not giving anyone enough power to have such a huge influence in selecting ALL top 20 witnesses basically alone.
Yeah, exactly, the intend is to centralize the system. Definitely. That's what Blockchain is about. Allowing huge stakeholders to control all the witnesses, that's not absolutely the reason why DPoS is so widely criticized. I don't want one person one vote, I want 10 votes. I stated that pretty clearly.
And no, I don't think it will result in splitting up stake. I don't think Pumpkin would split up his stake for this, nor do I think blocktrades would.
Of course this is centralization, centralization is having very few people having a lot of power over the system.
If that is a large stakeholder via a consensus node (PoS), a large stakeholder voting "his kind" into power (DPoS) or a large mining pool (PoW). All of these things undermine the principle of decentralization.
And as I mentioned PoW, all types of blockchain have similar issues.
If you'd read the literature, you'd know that all these blockchains are heavily criticized for these points. Because all of these blockchain allege decentralization, but in practice it's more a buzzword than real decentralization. (Minings pools, monopolization, etc)
And as I said, I trust a big pool of large stakeholders -> Yes
I trust 1-2 stakeholders -> No.
It is called single point of failure.
And I don't trust a single point of failure, if its a stakeholder or not.
I see there is no real point discussing with someone who calls the others person arguments "horseshit". Same kind of people who doubt man made climate change or think the earth is flat. What do I care about scientists, I can read a few blogs about the topic myself and I know things very well then.
These people who do distributed system research for many years, they know nothing.