Why complicate CIP-1694?

Here’s a proposal to make governance less complicated: an electoral college-type system.

I propose that the decision for a particular CIP will be coming from SPOs votes. The SPO vote will depend on the level of saturation, e.g., a saturated SPO vote is counted as 1 vote while a half-saturated SPO is counted as 0.5. The caveat is that the SPO vote will actually be decided by the SPO delegators, and the delegator’s vote will depend on their wallet holding. So, if 51% of the SPO delegation voted “yes” and 49% voted “no”, then the vote of the SPO is “yes” and must be counted as 1 if the SPO is fully saturated.

The pool operators, being technically savvy, will be tasked to polish the CIPs first before they are offered for referendum. Therefore, the CIPs must be acceptable to pool operators first before final referendum.

Is this strategy whale-resistant? Let’s take a look. If a whale has 34 M and concentrates that to a single SPO, that 34 M will generate only 1 vote. If the whale distributes that 34 M to 34 SPOs with 1 M each, that 1M will have a tiny influence, especially for a fully saturated pool. Further, the 1M will also have low influence if the whale distributes his stake to pools with low saturation because unsaturated SPOs votes will be less than 1.

Concrete sample calculation: If the whale plans to split his 34 M ADA to 3 SPOs with 10M delegation each (that means 11.33 M for each pool). Each pool will be able to have 0.31 vote [(11.33+10)/68]. Therefore, the whale will be able to garner 0.31x3 = 0.94 vote only if he splits his ADA into three pools.

Also, I don’t think we need DReps and the constitutional committee (CC). Dreps and CC are like senators and congressmen, and we know how that works.

3 Likes

Sounds feasible to me.

1 Like

So, what happens to SPOs who operate multiple pools ? Wouldnt they have a quite unfair advantage

1 Like

For multiple pools, each pool will have its own vote. The key here is that the pool’s delegators (not the pool operator) is what determines the vote of the pool which makes it decentralized. So for an MPO with 5 pools does not necessarily mean automatic all 5 “yes” or all 5 “no”. It could be mixed votes.

1 Like

First-past-the-post/winner-takes-all is already a horrible voting system in the anglo-american parliament and electoral college elections.

It gravely hinders the chances of third parties. (Which might not be relevant as long as we only have Yes/No votes.)

It can have outcomes, where the majority in the elected body is different than the majority in the popular vote. (Here, it would mean that the majority of stake pool votes would be No, although the majority of the stake voted Yes, but had a very unfortunate distribution on stake pools, losing a lot by narrow margins and wasting the major part on unnecessarily large wins.)

It is open to gerrymandering. Which is maybe less of a problem, since delegators can freely move, but maybe even more, since the alignment of delegators to stake pools is even more arbitrary than the alignment of voters to the constituency in which they live.

And there will probably be little indication beforehand, on which stake pool my vote might be promising.

First-past-the-post/winner-takes-all is already a horrible voting system in the anglo-american parliament and electoral college elections.

It gravely hinders the chances of third parties. (Which might not be relevant as long as we only have Yes/No votes.)

Unlike the US electoral process, we are not required by law to achieve 270 necessary electoral votes. This is the requirement that makes electoral college not ideal for a multiparty system.

If need be for multiple CIP candidates, we can do multiple referendums, each time eliminating the CIP that garners the least electoral vote.

It can have outcomes, where the majority in the elected body is different than the majority in the popular vote. (Here, it would mean that the majority of stake pool votes would be No, although the majority of the stake voted Yes, but had a very unfortunate distribution on stake pools, losing a lot by narrow margins and wasting the major part on unnecessarily large wins.)

If by popular vote you mean the “amount of ADA”, then that’s a worse metric. If the CIP garners 34 M ADA from a SINGLE person, that does not mean the CIP is popular. That single person just holds a lot of ADA, and this is the kind of impact that we are trying to minimize.

It is open to gerrymandering. Which is maybe less of a problem, since delegators can freely move, but maybe even more, since the alignment of delegators to stake pools is even more arbitrary than the alignment of voters to the constituency in which they live.

Assuming you control multiple pools, how do you actually do a gerrymandering among your pools? You don’t have the power to rearrange the stakes among your pools in such manner that will allow you to have your pools to vote in a certain way. You also do not have any information as to which way a wallet is going to vote. Real world gerrymandering needs this information.

And there will probably be little indication beforehand, on which stake pool my vote might be promising.

That “little” is better than having a whale exerting enormous influence via a popular vote. But then that “little” might not even exist.

Sorry, but I do not follow your logic.

  1. 3 SPO-s with 10M stake want the same (voting for the same thing)
  2. 1 whale with 34M stake wants the same as the 3 SPOs above (voting for the same thing)

Therefore, the 4 parties have the same influence independently of their own wishes.
Examples:
3 * 10/68 + 1 * 34/68 = 1*(10+34)/68 + 2 * 10/68 = 3 * (10+34/3)/68

2 Likes

Ah yeah. What the example wants to show is what happens when a whale tries to influence pool votes by delegating his ADA to the pools.

  1. So, the whale will have to split his delegation into three. That is, 34/3 = 11.33 M.
  2. the whale then delegates 11.33 M into each pool already containing 10 M stake. Total stake is now 21.33 M for each pool.
  3. Since the whale now has the majority of the stake, he controls the vote of each pool. But each pool vote is only 21.33/68 = 0.31 (68 M is the saturation level).
  4. Therefore, the total vote that the whale can accumulate from the three pools is 0.31x3 = 0.93.

It’s pretty interesting that even in your example you have the whale and the 3 pools (4 parties) voting independently, yet we arrived at the same number. Mine is just rounded off.

1 Like

No, because it’s compared to the whole voting, and not to the saturation level, what you did.

Meaning your calculation is relevant to the sat level part only of the voting and not to the whole voting.
It’s like absolute versus relative values, what the media like to use to fool the ppl. E.g.: The cancer risk is 50% higher (relative value), while absolute is (0.005% versus 0.0075%) as an example.

It’s a basic rational number addition. So, it does not matter whether it’s done your way or my way. It will end up being the same number.

Then you should understand where is the logic failure in your explanation.

Think of this again:

It does not matter if he splits to 3 SPOs (who want to vote to the same) or just put it into one SPO (who wants the same vote) the influence will be exactly the same.

4 Likes

It was a failure of communication. The phrase “0.31x3 = 0.94 vote only” meaning of his 34M ADA he is only able to muster less than 1 vote instead of a massive 34M votes.

But yeah, I did not see the possibility that you saw. The mechanism still does not guarantee a whale any massive influence.

Those 270 are just the majority. Of course, a vote on an action in CIP-1694 would also have to get a majority. And that’s also not the main problem with the system. The main problem is the majoritarian system below, where all votes that did not get the majority in a state (or in a constituency in the case of parliament elections) are effectively discarded.

Your system does not ensure at all that the undemocratic consequences of winner-takes-all harm whales. It could just as well harm the small fish, maybe even more.

Yes, we don’t have authorities shaping the constituencies. So, real gerrymandering is not possible.

Something very similar is possible. And it does not even need to be an SPO. A whale could distribute its stake among a lot of small pools, where it is in each case just enough to have a majority among the delegators to that pool. Your system says that the pool then votes with its whole delegated stake the way the whale wants. They get the additional voting power of the other delegators losing to them for free.

If I where a whale, I would specifically choose pools whose operators were very vocal on the other side than me – assuming that their usual delgators also are – to raise the chances that I am neutralising as much as possible of the opposition with this method.

1 Like

Those 270 are just the majority. Of course, a vote on an action in CIP-1694 would also have to get a majority. And that’s also not the main problem with the system. The main problem is the majoritarian system below, where all votes that did not get the majority in a state (or in a constituency in the case of parliament elections) are effectively discarded.

The problem with the popular vote is we don’t really have a popular vote since the influence of each unique vote is proportional to how much ADA the vote holds. We’ve seen many type of attacks in the Catalyst and in other blockchains exploiting the weaknesses of this “popular” referendum. The electoral vote appears to tamper that influence and might be more suited for a blockchain infrastructure.

Your system does not ensure at all that the undemocratic consequences of winner-takes-all harm whales. It could just as well harm the small fish, maybe even more.

Undemocratic consequences…such as? Popular vote works well if either you KYC or you have a decentralized identity solution.

A whale could distribute its stake among a lot of small pools, where it is in each case just enough to have a majority among the delegators to that pool. Your system says that the pool then votes with its whole delegated stake the way the whale wants. They get the additional voting power of the other delegators losing to them for free.

Yes, but remember that a pool vote is proportional to its delegation. You get 1 vote only if the pool you are in is fully saturated. When you delegate to small pools, the power of your vote also decreases. There is some influence but very minor compared to when you use a popular vote. Imagine a whale getting only 1 vote vs millions of votes (because he holds millions of ada).

If I where a whale, I would specifically choose pools whose operators were very vocal on the other side than me – assuming that their usual delgators also are – to raise the chances that I am neutralising as much as possible of the opposition with this method.

Well, just like any system there will be exploits, but it appears that under an electoral college system, a whale may have little influence to the outcome in comparison to the usual popular referendum.

The ones winner-takes-all systems are widely criticised for for decades? That your vote is probably always for the bin if you are a Republican in San Francisco or New York or if you are a Democrat in rural Texas?

On the one hand, one might argue that voting power by ADA, by skin in the game is a good thing, that there is no reason why someone with just a couple of ADA should have the same power as someone who has the risk of losing value on millions of ADA, that it would be too easy for malevolent actors to just buy into having voting rights on Cardano with 10 ADA.

On the other hand, yes, if you want to have something in the direction of “one person, one vote”, you need identity. And KYC or digital identities is not even enough. It’s ridiculously easy to create hundreds of digital identities and still possible to KYC lots of accounts. You need to somehow ensure that one person can only activate one voting account, you need uniqueness, you need enough information to rule out multiple registrations per person. And that is a lot of information.

One person, one vote is probably not possible in a decentralised manner at all.

It’ just a constant factor of 1/saturation. It doesn’t matter at all if you say all pools have as many votes as ADA delegated to them (millions each) or if you say they all have delegated ADA divided by saturation (at most one each).

Voters with smaller bags also only get an even more tiny fraction of one vote in your system and are, of course, also affected by the pool they are voting in only getting a fractional vote if it is not saturated, but they at least have hundreds or thousands (just not millions like the whales) of votes in a direct vote by ADA.


Simplified example:

Popular vote by ADA would be: 99 million yes, 97 million no
Yes wins with 50.5 % of the votes.

Now, your “pool electoral college” is used with 5 pools:
Pool 1: 50 million yes, 20 million no, 70 million of 70 million saturation, 1,0 votes yes
Pool 2: 26 million yes, 2 million no, 28 million of 70 million saturation, 0.4 votes yes
Pool 3: 12 million yes, 2 million no, 14 million of 70 million saturation, 0.2 votes yes
Pool 4: 65 million no, 5 million yes, 70 million of 70 million saturation, 1.0 votes no
Pool 5: 8 million no, 6 million yes, 14 million of 70 million saturation, 0.2 votes no

Yes wins with 1.6 against 1.2 (57.1 %) of the votes.
This distribution makes the Yes win much clearer than it really is.

But 63 million of the no in Pool 4 are a whale and that whale is smart. They redistribute their stake so that they have a majority in Pools 2 to 5 (28 million to Pool 2, 14 million to Pool 3, 7 million stay in Pool 4, 14 million to Pool 5).
And that strategy they can even do without knowing or guessing how the other delegators will vote, because they then have 50% themselves in all those pools.

Pool 1: 50 million yes, 20 million no, 70 million of 70 million saturation, 1,0 votes yes
Pool 2: 30 million no (28 from whale), 26 million yes, 56 million of 70 million saturation, 0.8 votes no
Pool 3: 16 million no (14 from whale), 12 million yes, 28 million of 70 million saturation, 0.4 votes no
Pool 4: 9 million no (7 from whale), 5 million yes, 14 million of 70 million saturation, 0.2 votes no
Pool 5: 22 million no (14 from whale), 6 million yes, 28 million of 70 million saturation, 0.4 votes no

Now, no wins with 1.8 against 1.0 (64.3 %) of the votes.
Yep, the pool the whale was originally delegated to went from 1.0 voting weight down to 0.2, but securing (without doubt) the majority in three other pools more than compensated that.
A single whale (with 32.1 % of the total ADA in this toy example) had the possibility to control the voting of 4 of our 5 pools and swing the vote.

3 Likes

I will only answer your math argument because the discourse is getting convoluted.

What you have is a good point.

Now imagine that, like the whale, retails are also smart and already KNOW that when a pool is 50% whale, that pool is CENTRALIZED and that their vote is irrelevant. So, they find someplace else that is more DECENTRALIZED. That leaves the pool where the whale is with lower vote count than he expected. There is a dynamic here - a game theory - just like how the consensus mechanism (although flawed) was designed. In a popular vote, your power is only dependent on how much ADA you hold. Nothing else.

Now we can argue that retail, being a representative of a population, only really have an average IQ which brings me to my other idea (read it here) that pool reward, block production, and now voting power, should be proportional to how decentralized the pool is.

There is probably already an impossibility proof for this published somewhere… Nevertheless:

I will contend that it is impossible to develop some game theory mechanism that distributes voting power disproportionately to small holders, in the setting where you are unable to prove the identity of every voting wallet.

Are you suggesting that Cardano KYC every wallet in order to vote? If not then whales will just split their wallets in the most optimal way possible to maximise their voting power in order to overcome any system design that seeks to somehow reduce their voting power. Furthermore, whales are very likely to be more sophisticated at doing so than the average small holder.

All this cardano-whale bashing is getting too much. No wonder he closed his twitter account. Sorry - bad joke.

1 Like

I will contend that it is impossible to develop some game theory mechanism that distributes voting power disproportionately to small holders, in the setting where you are unable to prove the identity of every voting wallet.

Game theory model can be developed for any system where there are competing players. I don’t think the game theory provided any exemptions, and I don’t see how the proposed voting system within a blockchain infra is exempted. Also, just like the consensus mechanism, the voting game theory model need not know your identity. Although, I have to admit I am not a mathematician, but I’d like to believe I am decent at math.

Are you suggesting that Cardano KYC every wallet in order to vote?

Not really. I mentioned KYC in relation to the popular referendum.

If not then whales will just split their wallets in the most optimal way possible to maximise their voting power in order to overcome any system design that seeks to somehow reduce their voting power. Furthermore, whales are very likely to be more sophisticated at doing so than the average small holder.

Empowering decentralized pools like my proposed mechanism that I described here solves this. Why? Because it is highly likely that in a decentralized pool, there are more small investors than there are more whales. Therefore, the pool vote stems from delegator consensus rather than from big player influence.

All this cardano-whale bashing is getting too much. No wonder he closed his twitter account. Sorry - bad joke.

That was not too bad actually. LOL.

The problem is that there is no way to tell the difference between a million small investors and a whale that has split his wallet into a million “small investor” wallets.

1 Like

If the whale splits his wallets such that for each wallet delegated to a pool has a holding of < 50% of the pool holding, then he dilutes his influence over each pool. If it is >50%, then pool decentralization metric decreases and small investors leaves the pool, decreasing the pool’s vote count.