Jun 8, 2023 | Voltaire era: Parameter committee intermediate state

[Cristina_Echeverry]

The Parameter Committee (interim state) meets on a bi-weekly basis and as of June 8th it will be holding meetings on a 3 week basis. Its members are a mix of different Cardano community members. The Parameter Committee discusses all parameters relating to the Cardano protocol including network, technical and economic parameter type. They are a technical advisory group and only make recommendations.

Updated list of the Members (interim state):

Chairs:
Chair: Kevin Hammond
Vice- chair: Alex Moser
Vice- chair: Vijay Bhuvangiri

Advisory Group Heads:
Network group: Neil Davies
Economic group: Samuel Leathers
Technical group: Markus Gufler

Advisory Group Members:
Network group: Karl Knutson, Matthias Sieber, Marcin Szamotulski
Technical group: Ruslan Dudin, Michael Peyton Jones
Economic group: Sergio Sanchez, Philip Lazos, Giovanni Gargiulo

Other:
Communications: Tommy Kammerer, Nathaniel Acton, Addie Girouard
Domain experts: Martin Lang, Andrew Westberg
Observers: WIP
Secretary: Cristina Echeverry

Agenda and Updates:

Agenda

• Reminder of meeting recording
• Discuss the SPO poll results (PCP 001)
• Cristina to create each advisory groups discord channels

Updates

• The main discussion topic has been the SPO poll, results are under interpretation, however here is a brief summary of what was discussed:

  • Asked questions:
    • Which (k and min fee) setup would you prefer to be put in place from Q3 2023 onwards?
      • Keep k at 500 and minPoolCost at 340 ada
      • Keep k at 500 and halve minPoolCost to 170 ada
      • Increase k to 1000 and keep minPoolCost at 340 ada
      • Increase k to 1000 and halve minPoolCost to 170 ada
      • I would prefer to abstain
      • None of the provided options

• Preliminary results:

• Total SPO votes: 797

Keep k at 500 and minPoolCost at 340 ada	8.78 %
Keep k at 500 and halve minPoolCost to 170 ada	12.30 %
Increase k to 1000 and keep minPoolCost at 340 ada	23.84 %
Increase k to 1000 and halve minPoolCost to 170 ada	45.80 %
I would prefer to abstain	0.63 %
None of the provided options	8.66 %

• Total participated stake: 10.87b ada

Keep k at 500 and minPoolCost at 340 ada	2.08b
Keep k at 500 and halve minPoolCost to 170 ada	2.73b
Increase k to 1000 and keep minPoolCost at 340 ada	1.47b
Increase k to 1000 and halve minPoolCost to 170 ada	3.00b
I would prefer to abstain	79.56m
None of the provided options	1.51b

• There was a variety of SPO voters. If we see the different filters the results mainly point to Increase k to 1000 and halve minPoolCost to 170 ada, however the results are not conclusive - especially for keeping or increasing k.

• Next steps:

  • Results are now under interpretation by the parameter committee and the Cardano community.
  • The advisory groups will have to consider the poll outcome from the perspective of their parameters. Sam Leathers and his group will consider this from an economic perspective, Markus and his group from a technical perspective and Neil from a networking perspective. Then they will come back with recommendations at the next meeting, and the whole committee can come to an official recommendation.

• Sources: SPO Polls 96861fe7da8d45ba5db95071ed3889ed1412929f33610636c072a4b5ab550211 - Cardanoscan

• Cardano Explorer - AdaStat

Action items

• Cristina to confirm all the advisory groups discord channels
• Advisory groups to revise the PCP 001 and provide a recommendation

[COSDpool]

It would have been an important realisation from the vote review that some SPOs included metadata comments with their votes. This metadata commenting was pointed out to committee members while the vote was taking place, so not including it in the summary above sends a message that these comments will be officially ignored now & in the future.

One reported example, including why some SPOs felt the need to include these metadata comments attached to their quantitative vote:

Entering Voltaire: poll experiment live on mainnet

Especially for a poll with impoverished options, it’s constructive to add a Tx metadata message, particularly when voting “none of the above” to explain why and/or what missing choice one would have selected. I’ve also seen cases where this would make sense even with the canonical options (when not fitting the voter’s real desire), or to explain why the voter is abstaining. Submitting the additional metadata with the voting transaction according to CIP-0020 will produce a readable message alon…

The option to submit metadata with the linear voting was considered important enough for the CIP-0094 architects to include it in the CIP itself (upon demand):

• https://github.com/cardano-foundation/CIPs/pull/496#pullrequestreview-1465358443
• https://github.com/cardano-foundation/CIPs/pull/496/commits/3c2e5b74da534fba4876d74bfe37bca23592f8f3
• https://github.com/cardano-foundation/CIPs/blob/master/CIP-0094/README.md#adding-context

[Kevin_Hammond]

Thank you for pointing this out. We are, of course, very interested in the commentary and contextual information that SPOs have provided around this poll. It will help inform both this and future discussions on parameters: CIP-1694 explicitly includes metadata hashes on votes as well as actions precisely so that explanations/evidence can be provided. I’ll discuss with the CF how this can be collected - there is no automated tooling for this yet, of course.

[ccgarant]

@Cristina_Echeverry

however the results are not conclusive - especially for keeping or increasing k.

I mean… it’s Proof-of-Stake, the winning vote by % stake should win. That’s a core, promised utility of $Ada.

[werkof]

I agree and support the idea that the comments-metadata will be crawled and part of the report. But there will not be any interpretation like sumarizing “7 operators posted comments for 9 pools, basically saying X, while other 15 comments were more in favour of Y”

Just to clarify: This summary of the last meeting reflects what was actually discussed there. At the time of this meeting, there was no report at all that included the metadata comments. Based on this initial reflection, the subcommittees should begin to look at the survey results. In case someone wanted to look at additional comments, they are perfectly visible in the public Dashboards.

However, since there were a decidedly large number of vacations and also other tasks (e.g., travel to Voltaire workshops), it was difficult to define group meetings.

There is definitely some progress, but it still needs some time. The next of the three-week PC meetings is already coming up…

[ccgarant]

@Cristina_Echeverry

The advisory groups will have to consider the poll outcome from the perspective of their parameters.

Expectations are the polling results are made public in these notes like FOMC rate votes. Maybe also skin-in-the-game stake / pool disclosures too.

[Cristina_Echeverry]

As mentioned above by Werkof:
Just to clarify: This summary of the last meeting reflects what was actually discussed there. At the time of this meeting, there was no report at all that included the metadata comments. Based on this initial reflection, the subcommittees should begin to look at the survey results. In case someone wanted to look at additional comments, they are perfectly visible in the public Dashboards.

[adatainment]

Hello,

Thank you for posting the notes @Cristina_Echeverry.

For someone who finds the post in the future and wonders what percentage participation that was: worth mentioning that the “total participated stake of 10.87b ada” represents a participation rate of 47.81%.

(based on the 22.738b ada delegated to stake pools at the end of the delegation phase)

[Homer_J]

For next Parameter committee meeting (or before it is fine too) could we please get an official statement from IOHK on the subject of what has happened to the research team’s findings that were referenced in March 2021 blog by Colin Edwards (https://iohk.io/en/blog/posts/2021/03/04/not-long-till-d-0-day/), which stated:

“Our research team is in the late stages of finalizing an approach. They will soon present their findings, and we will update the community then. However, the team has now concluded that a0 should change. We believe this change will greatly benefit the network, making the system more sustainable, widely distributed, and globally diverse.”

Thanks.

[girljoseph]

@Kevin_Hammond, it’s alarming that YOU are the Chair this committee.

Shortly before announcing the formation of the parameter committee you commissioned a blog post that claims:
“Minimum fees act as an additional layer of protection against Sybil attacks {…} Abolishing or lowering these minimum fees would remove this protective layer and could leave Cardano more open to certain forms of economic attack.”

• Blog > 2022 > October > Staking parameters and network optimization – where next for k and min fee?

This same message was also repeated as a prepared statement in more than one SPO call.

To date there has not been a rational or scientific argument backing that statement. In fact, all of the researched published by IOG and citizen scientists directly opposes that claim. Everything I’m finding is pointing to YOU as the source of that statement. So I’m hoping you would be willing to defend it.

The peer-reviewed reward sharing scheme and the design rationale explains sybil protection is derived from a real cost of pool deposit and the ability of pools with higher pledge to offer higher rewards. Splitting pledge should result in lower utility/desirability.

minPoolCost makes it so that an established pool with 0 pledge has the ability to offer higher rewards than a new pool with the HIGHEST public pledge on the market. This is one of many examples of how minPoolCost is opposing researched rational ideas. I would be happy to share additional problems minPoolCost creates.

Can you please rationally and scientifically defend your point about minPoolCost not only providing sybil defense, but also how it can provide ADDITIONAL defense while it simultaneously destroys the notion and power of pledge?

[earncoinpool]

@Cristina_Echeverry

however the results are not conclusive - especially for keeping or increasing k.

Can you explain how the committee can claim the results are NOT conclusive? AdaStat show the vote using 5 metrics. K=1,000 and min fee of 170 won the poll using each of the 5 metrics.

[werkof]

In my opinion the only effective meaningful and important metric is stake. Not head counting, popularity or whatever else. People who have signed up for Cardano and PoS signed up for that, not any other “on-demand” replaced metric.

By looking at the stake results grouped by the 4 combined answers, there was a surprisingly balanced situation between two options at 27.6 vs 25.1% of support (1).

By grouping the answers by parameters (you “forgot” this in your screenshots ;- ) the results are different:
44.3% supporting keep k as is at 500 vs
41.0% supporting k increased to 1000. (3)

The support for halving minPoolCost is clear in both views. (2)

Hence the statement and the widespread opinion in the parameter committee: “…the results are not conclusive - especially for keeping or increasing k”

So here, based on the surprisingly close result, a situation was created that actually enables exactly what was one of the main goals of this first poll: to recognize, based on practical hands-on experiences, which things in the design of CIP-1694 need to be clarified and defined more precisely, e.g. what kind quorum and clarity is needed in a voting process.

If you are interested in these topic and contribute with opinions and ideas, there are others who already did, eg @HeptaSean at

Minimal Viable Governance

Before anyone asks: I am not aware that this has ever been proposed or used somewhere, unfortunately. Quora seem to be set rather ad-hoc almost everywhere, even in quite important “official” rules. I’d be glad if someone finds in-depth sources. For comparison, these are the alternatives often found: [ParticipationQuorum][MajorityQuorum][MarginQuorum] Even the – objectively bad – participation quorum is not only used in Catalyst, but also in the rules for state- or nation-level referenda. C…

[girljoseph]

Considering this was an SPO poll, stake in the system - aka pledge, seems the metric that matters.

We know very well that stake is tremendously sticky. We even have a parameter, minPoolCost, that incentives it to stay put.

Would you be willing to share the amount of stake that redelegated within the relegation window compared to avg stake movement?

[werkof]

Indeed that chart is one part of the report, I currently work on.

As you can imagine/ as it was already visible in the dashboards, the redelegation was not overwhelming.

Now a question to you: As my chart in the report would be a single point of truth, could I gently ask You, to also prepare such a visualization for the community? I believe we should have at least two sources for this, and it looks to me like you are interested and capable in this topic.

[girljoseph]

Without a single argument or shred of science in favor of a minPoolCost the entire poll is pointless, IMO. minPoolCost to 0, the only scientific choice, wasn’t even an option… This poll is effective at identifying SPO’s who don’t really understand the network’s design or those maximizing value extraction.

I’ve spent a lot of time already creating visualizations and educational materials for the community (or anybody interested) about minPoolCost and its effects, which are all damaging. This material was directly undermined by IOG in preparation for this poll. I’m not going to waste my time to have Kevin heavy-hand a decision like he’s done for the last 3 years.

[werkof]

Sorry, no! Do not try to escape now

You asked about delegation movements. A new thing in Cardano with this survey. You definitely did not chart this in the past, and I politely asked you to provide an alternative, confirming or contradicting visualization of how much stake moved in the two redelegation epochs using on-chain data.

This has nothing to do with the two parameters of the poll, but with the concept of the poll.

You made (imo) a pointless claim against the mPC you so hate, that it would even prevent redelegations, and tried to validate your own claim by linking it to the polls redelegation phase.

Ok, then I’m happy to leave you with that opinion and disagree. But then let’s at least do something productive and useful: Would you please compile some data and show the ecosystem how much or how little the shares have moved during the poll epochs?

So many here are calling for decentralization, distribution of power and decision making. And at the same moment they are asking for a report or justification from someone else. Or for someone else to change their mind because there is no doubt about the one and only acceptable truth.

Show that you can make a constructive contribution. I believe in you!

[girljoseph]

I conducted an analysis about 7 months ago on redelegation trends. Here were my network wide findings.

We can see not much stake is moving around, but we’ve all known that anecdotally already. Asking for redelegation as a vote was maybe a test to see if people would, but expecting them to was unfounded. If about the same stake moved around during the redelegation phase as would move around on average anyway, then using stake as vote weight is silly. Almost intentionally silly. The network has spent 3 years ossifying stake into old pools.

Do we not believe in incentives anymore? By asking delegators to redelegate as a vote you are asking them to potentially accept lower rewards on top of a transaction fees to and from that vote pool. That’s on top of understanding the parameters themselves, which IOG made a point to misinform people in preparation for this poll. We’re asking people to pay to vote about things they don’t even understand.

[girljoseph]

I didn’t intend to use the lack of stake movement during the redelegation phase as proof minPoolCsot makes stake sticky.

minPoolCost makes stake sticky because - who wants to accept guaranteed lower rewards by changing pools? The network incentivizes people to remain in their high stake pool instead of choosing a pool that would offer better rewards at saturation by having higher pledge and higher performance.

[werkof]

Many large and mid-sized pools chose all available options. Those who wanted could easily find an alternative pool with the preferred answer without losing rewards.

One question that interests me is how we can deal with our governance and best shape it: Would you be willing to change your mind or accept a different opinion?

[werkof]

@girljoseph

Asking for redelegation as a vote was maybe a test to see if people would, but expecting them to was unfounded.

hmmm maybe this is one of the reasons why we can’t find a common denominator.
Do we have to or do we want to assume that Stake has to move? Or in other words, that the majority of people are dissatisfied and demand a significant change? Or does it just seem to some that this should happen because they themselves want a change and can’t understand why a broad majority doesn’t support it?

@girljoseph

If about the same stake moved around during the redelegation phase as would move around on average anyway, then using stake as vote weight is silly

Pardon, this claim sounds pretty silly!
Such a claim would only make sense if you want to deeply and firmly assume that people are dissatisfied and want a change. But this statement excludes that people are simply satisfied with what they get in rewards and to whom they have delegated.

I think that the pool landscape and the quite diverse strategies and activities of operators over the years has also led to delegators landing where they feel confident and comfortable. For example, there are operators who have created dedicated chat groups for their pools and communicate with them on an ongoing basis. They have then made polls in advance on the occasion of this poll. Delegators who appreciate this will be convinced of this pool and will also trust their votes. Other pools have attracted delegations with 0% margin, or with Youtube videos etc etc.

The possibility for delegators to redelegate is a possibility, but definitely not an obligation. If they didn’t do it, it’s because they agree with their pool’s opinion, or have no interest in it, or didn’t even realize it. Do you really want to make a claim, stake didn’t move that much because delegators fear to get less rewards?

In the end, this is a nature of delegated PoS. The combined stake increases the chances of generating blocks.

And also regarding Pledge I can add my opinion here: Personally, I believe that Pledge - with whatever value and in whatever linear or creative curve - can at best benefit as much as harm at the same time. But maybe someone will come up with a really good suggestion, something like a 4d version of a square sphere.
Pledge in its current limited effectiveness has contributed to stake distribution. That stake distribution can create blocks and - more recently - participate in polls. So I really see no reason to give this Pledge factor any special importance in voting results again. The only one would be maybe to give Pledge more importance only for votes but not for blocks and rewards.

minPoolCost was then, at the time when you finally had to set the parameters to launch the mainnet, a way to avoid sybill attacks, because you had strongly mitigated a0 - also concerning its protection factor for sybill attacks.

what remains? 3 questions:
Is there a perfect world?
Is it conceivable to live and stay in a less than perfect environment?
Or is the solution simply that everyone only needs to follow me into my idea of a perfect world, and then everything is fine?