Flare/Songbird vote power cap vs. Cardanos's "pool saturation"

[Gringo]

Cardano uses the concept "pool saturation" when a validating pool is staked above 64 M ADA while currently the Flare/Songbird ecosystem uses the vote power cap concept with a vote power cap currently set at 2.5%.

While on Flare it is expected one FTSO per "real" team on Cardano this is not a problem. One single validator usuallty offers more than a "validating" pool so that when one pool is saturated (more than 64 M ADA staked), users can chose another pool WITHOUT changing the validating team doing the work. I think this brings more transparency, as there is less incentives for collusion.

On Flare/Songbird we have witnessed collusion aimed at controlling more than the 10% power cap and since now the vote power cap is being reduced to 2.5% it is posible that the incentive for collusion is greater (4 times greater).

We should understand that what is important to the network is the accuracy of data, so an FTSO provider with a 89.66% accuracy should be better than one with a 27.36% accuracy. So it should get MORE rewards. But it is not the case, because of the power cap set a 2.5%

So now we are at this situation:

A-FTSO from Alex Dupre has been in general around a 10% vote power and that's simple because he's done a hell of a good job validating. Since it is still at 8.43% vote power cap suffers a lot of "rewards penalties" allthough the accuracy numbers (89.66%) are top notch.

Maybe is something of a different culture on the Flare network (regarding Cardano), but what if he provided this service:

A-FTSO1

A-FTSO2

A-FTSO3

A-FTSO4

so its up to users to chose if they want to delegate to the same entity (running 4 FTSO's) in a transparent way, without it being consider "collusion".

I suspect there still are teams of 4 or more FTSO colluding and therefore controlling more than 10% power cap, but in the end is "good actors" like A-FTSO that are being penalized ironically because of the collusion problem.

A good place to check real time stats of FTSO's to understand this problem is:

                              https://app.blazeswap.xyz/stats

 

 

[BillyOckham]

3 hours ago, Gringo said:

Cardano uses the concept "pool saturation" when a validating pool is staked above 64 M ADA while currently the Flare/Songbird ecosystem uses the vote power cap concept with a vote power cap currently set at 2.5%.

While on Flare it is expected one FTSO per "real" team on Cardano this is not a problem. One single validator usuallty offers more than a "validating" pool so that when one pool is saturated (more than 64 M ADA staked), users can chose another pool WITHOUT changing the validating team doing the work. I think this brings more transparency, as there is less incentives for collusion.

On Flare/Songbird we have witnessed collusion aimed at controlling more than the 10% power cap and since now the vote power cap is being reduced to 2.5% it is posible that the incentive for collusion is greater (4 times greater).

We should understand that what is important to the network is the accuracy of data, so an FTSO provider with a 89.66% accuracy should be better than one with a 27.36% accuracy. So it should get MORE rewards. But it is not the case, because of the power cap set a 2.5%

So now we are at this situation:

A-FTSO from Alex Dupre has been in general around a 10% vote power and that's simple because he's done a hell of a good job validating. Since it is still at 8.43% vote power cap suffers a lot of "rewards penalties" allthough the accuracy numbers (89.66%) are top notch.

Maybe is something of a different culture on the Flare network (regarding Cardano), but what if he provided this service:

A-FTSO1

A-FTSO2

A-FTSO3

A-FTSO4

so its up to users to chose if they want to delegate to the same entity (running 4 FTSO's) in a transparent way, without it being consider "collusion".

I suspect there still are teams of 4 or more FTSO colluding and therefore controlling more than 10% power cap, but in the end is "good actors" like A-FTSO that are being penalized ironically because of the collusion problem.

A good place to check real time stats of FTSO's to understand this problem is:

                              https://app.blazeswap.xyz/stats

 

 

I think you make some good points here.  This clearly has been a problem, and it’s unclear to me how the smaller cap is going to affect it.

From a few years interactions I am very confident that @ftso_au on here is a good actor in this space and also a reliable network supporter.

So I wonder if (s)he has any opinion on this issue?

 

 

[BillyOckham]

Having said the above….   in fact, given the timing, I doubt they will be able to reply because of being frantically busy and this will probably pass unnoticed by them.  Oh well, maybe revisit in a week or two.

 

[ftso_au]

5 hours ago, BillyOckham said:

Having said the above….   in fact, given the timing, I doubt they will be able to reply because of being frantically busy and this will probably pass unnoticed by them.  Oh well, maybe revisit in a week or two.

 

There’s lots of internal discussion regarding this issue, and there’s a few short term solutions that will make up part of STP-02.

Longer term, it’s most likely the FTSO will move to a staking model, at-least partial staking, which will include slashing.

And yes @BillyOckham I’d like to think we set the bar for ethical behavior in the ecosystem! 

[Gringo]

Bifrost suggestion just means that they are framed on the XRPL mentality. They just want to recreate the UNL (Unique Node List) concept of XRPL on Flare.

The point is, XRPL doesn't pay rewards, so for the end user it doesn't matter who validates the network. Flare Network is run by economic incentives and asking the end user to avoid optimizing their rewards on a context of a high inflation network is futile.

Bifrost positions itself as a judge ruling which FTSO is "misbehaving" and which not. But they are also an FTSO provider. They can't be a judge for the network and also an FTSO (that's "misbehaving" as well!).

Because of this particular aproach, they hide certain FTSO for their users, so they are in fact trying to create a sort of UNL on Flare.

Conclusion: The 2.5% vote power cap hasn't solved anything.

Edited January 15 by Gringo