Jump to content

FTSO matters


BillyOckham
 Share

Recommended Posts

13 hours ago, Seoulite said:

I wonder if this is still the plan or not, considering when you delegate you are changing all your FLR to WFLR. 

I don't think it will be all or nothing, you'll be able to choose how much to deposit into and withdraw from the voting smart contract.

Link to comment
Share on other sites

19 minutes ago, brianwalden said:

What's different from the whitepaper is all of our expectations and interpretations while we wait to see the actual implementation.

Well said. I too, often go back to the whitepaper and it all seems to line up. Looking forward to more information released in the near future.

Link to comment
Share on other sites

  • 1 month later...
On 6/7/2021 at 9:24 PM, FTSO_AU said:

We’ve started the FTSO AU Discord, I welcome you to join us over there … https://discord.com/invite/y6dc6AsTGx

I’ll still, of course, be active here.

Hi, can I ask you about the price feeds please?

I notice at https://www.ftso.com.au/#price_signals that the price listed for BTC seems a fair way away from the Binance BTCUSDT price I generally use.  Significantly lower:

Binance:  $38925.       FTSO $38126

Can you explain a bit about that please?

Link to comment
Share on other sites

I have few questions in my head.

 

On 5/27/2021 at 10:46 AM, FTSO_AU said:

Q./ How does the FTSO determine if the prices the FTSO providers are providing are accurate? I understand the top and bottom 25% votes are discarded, but how does the FTSO know the middle 50% are accurate?

A./ It’s based on consensus. All Flare networks users can vote with Spark or the corresponding Fasset, so the price assumed to be accurate if a majority of people believe it to be.

Let's assume there is a Data provider which does extremely good and he earns constantly the reward (Spark).

Isn't it expected that this will draw attention  to Spark holders, making them delegate their vote to that Data provider? 

1) Doesn't that mean that the voting will get centralized over time since more and more people will use the same Data provider? 

2) If one data provider gets like 70% of totals votes, how can 30% outvote him?

3) Can't that one data provider intentionally provide wrong data and get benefit out from it if he has like 80% voting power? For instance, can't he say that XRP price is 0.0001 dollar, making flare loan / mutual go mental and people lose money?

@brianwalden 

@Flintstone

guys, what is your opinion here? I would be happy to hear it.

Thank you

Edited by Kiwi
typos
Link to comment
Share on other sites

3 hours ago, Kiwi said:

I have few questions in my head.

 

Let's assume there is a Data provider which does extremely good and he earns constantly the reward (Spark).

Isn't it expected that this will draw attention  to Spark holders, making them delegate their vote to that Data provider? 

1) Doesn't that mean that the voting will get centralized over time since more and more people will use the same Data provider? 

2) If one data provider gets like 70% of totals votes, how can 30% outvote him?

3) Can't that one data provider intentionally provide wrong data and get benefit out from it if he has like 80% voting power? For instance, can't he say that XRP price is 0.0001 dollar, making flare loan / mutual go mental and people lose money?

@brianwalden 

@Flintstone

guys, what is your opinion here? I would be happy to hear it.

Thank you

I've thought about this too. I believe the oracle throws out the outliers, so they couldn't literally crash the price like that. But I'm sure that if someone got the point where they have that much of the voting weight they could at least nudge the price a little bit. A penny or two can make a big deal to arbitragers.

The obvious comparison is to staking. I stake BNB on Binance and there doesn't seem to be any real rhyme or reason to who wins more rewards than the others. One day my validator might be near the top, the next it will drop. My guess is that as long as the randomness of price feeds outweighs a data provider's ability to be consistently more accurate than everyone else, then no one provider will be able to dominate the others.

But there does seem to be some kind of perverse incentives. Ordinary people like us don't care about who's got the most accurate feeds, we're all going to vote based on who pays out the most. Data providers technically aren't rewarded for having the "right" price, no such thing exists. They're rewarded for being closest to the average. What if they started submitting the price they thought would be closest to the average rather than their real price prediction. They might be rewarded more, but they wouldn't be submitting the coin's actual price, just the price that they think will win. Or even worse, what if a data provider decided to wing it and just submit the last settled price from the last feed as their next prediction. I bet it would be close enough to win something some of the time, and they wouldn't have to do any work trying to come up with an accurate price.

My guess is that Flare has a much better understanding of this than I do and has thought about these things. Also, this is a very good reason for launching Songbird. This is the kind of thing that you can't test by just making sure the code works. You can only really test it when the system is really running and people are competing to make money against each other. Will the price remain consistent with other established feeds, or will it veer off like a stablecoin that loses its peg?

One more thing, I just learned about Terra. They're a PoS network, but they make their validators submit price feeds as well as network transactions. I don't really know the details of how it all works, but it may be worth looking into just to get an understanding of a similar system that's already running.

Edited by brianwalden
Link to comment
Share on other sites

1 hour ago, brianwalden said:

They're rewarded for being closest to the average. What if they started submitting the price they thought would be closest to the average rather than their real price prediction. They might be rewarded more, but they wouldn't be submitting the coin's actual price, just the price that they think will win

^this x 100

1 hour ago, brianwalden said:

My guess is that Flare has a much better understanding of this than I do and has thought about these things.

No. I believe you are wrong here. Flare have demonstrated over the last few months that they are not infallible and the choice of algorithm for rewarding oracles is one more example of a lack of thorough understanding of all the complications that a system like this brings into play. Just hoping that everyone will play fair is wishful thinking.

1 hour ago, brianwalden said:

Also, this is a very good reason for launching Songbird

Absolutely. If the system can be gamed, it will be gamed. All the oracles voting slightly higher or lower than the "true price" and then scalping joe public one way or the other is not beyond possibility. Look at banks and LIBOR rates for example. Since crypto is unregulated, it's almost compulsory for this to happen. 

Edited by jbjnr
EDIT: If it turns out that I'm mistaken and in fact the flare oracle reward algorithm is ok and I've misunderstood, then flare, please accept my apologies for criticising it now!
Link to comment
Share on other sites

Thanks guys for an interesting discussion.  I have been thinking along similar lines to much of the above.  So great minds think alike.

 

Or foolish ones seldom differ.      :) 

 

On a follow up to my observation about the https://www.ftso.com.au/#price_signals …

 

Now that the prices are slumping the price signals are higher than reality.
eg 72.4 cents XRP when reality was 70.4.  That’s a big difference.

 So it looks like a time delay issue to me.  Sorry to put you on the spot @FTSO_AU but can you explain this?

Link to comment
Share on other sites

11 hours ago, Kiwi said:

1) Doesn't that mean that the voting will get centralized over time since more and more people will use the same Data provider? 

2) If one data provider gets like 70% of totals votes, how can 30% outvote him?

It’s possible there’ll be a limit placed on how much voting power any one Signal Provider can have to avoid this situation.

 

1 hour ago, BillyOckham said:

So it looks like a time delay issue to me.  Sorry to put you on the spot @FTSO_AU but can you explain this?

Looking into this now. We fixed an issue a few days back with the aggregation function that we thought resolved this, perhaps not?!

There’s a delay we have on our pricing to avoid people simple scraping ours but this seems beyond that. 

I’ll update you soon. 

 

Link to comment
Share on other sites

28 minutes ago, FTSO_AU said:

It’s possible there’ll be a limit placed on how much voting power any one Signal Provider can have to avoid this situation.

 

Looking into this now. We fixed an issue a few days back with the aggregation function that we thought resolved this, perhaps not?!

There’s a delay we have on our pricing to avoid people simple scraping ours but this seems beyond that. 

I’ll update you soon. 

 

This new Canary Network idea (at least new to me) is looking very good about now eh!  
 

Resolving these type of System level interaction issues in a smaller pathfinding Canary network really seems like a good idea.

Having said that….  this issue popped up, and presumably will be resolved,  simply because FTSO_AU are pro-active and decided to have their signal out in advance of the release.  Kudos to you folk.

Link to comment
Share on other sites

6 hours ago, BillyOckham said:

This new Canary Network idea (at least new to me) is looking very good about now eh!

It absolutely is. We’re as keen as anyone to see Flare launched but we’re 1000% behind Songbird, it’ll make our service so much more robust I’m sure of it. Which, we’ve deployed an update to our ticker, so the pricing should look better. There’s still some work to be done until I’ll say it’s “fixed” but we know a workaround.

 

6 hours ago, BillyOckham said:

Having said that….  this issue popped up, and presumably will be resolved,  simply because FTSO_AU are pro-active and decided to have their signal out in advance of the release.  Kudos to you folk.

Thanks for recognizing this, that was our intention. We’ll soon be asking for registrations on our app, for the exact same reasons. Having fresh eyes on it and using it real world will be valuable.

I’ll let you know @BillyOckham would be good to have a few of you guys, including @brianwaldenfeedback.

Link to comment
Share on other sites

5 minutes ago, FTSO_AU said:

Which, we’ve deployed an update to our ticker, so the pricing should look better.

Yep sure does.  
 

I like how open you are to saying “it’s just a workaround for the moment” but that you will fix it.  That kind of straightforward lack of bs goes a long way with me.

I know that Hugo has said he wants people to support more than one FTSO provider,  but as things stand, the bulk of my support will go to you.  Unfortunately it won’t amount to much,  so don’t get too excited.  :) 

I like that you are Aussies too….  literally anyone from anywhere could be doing this stuff, but here you are… so supporting my fellow Aussies seems like a no brainer so far.    And I’m particularly well suited to no-brain operations.  :) 

Link to comment
Share on other sites

1 hour ago, BillyOckham said:

I like how open you are to saying “it’s just a workaround for the moment” but that you will fix it.  That kind of straightforward lack of bs goes a long way with me.

I’d lose credibility if I said it was fixed then you checked again at 9pm tonight and saw it had “drifted” again. 

We’re appreciative of your support, whatever the allocation. We won’t let you down! 

Link to comment
Share on other sites

 Share


×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.