dude, honestly. Duncan has actually spoken about this and made it pretty clear, quite clear in the podcast episode - i suggest you watch it diligently and rewatch it with an open mind (not with the assumption or preconceived notion that the system architects are out to screw the small pools)as perhaps some of the points that are not being picked up will crystallise.
2 Likes
dude, honestly. Duncan has actually spoke about this and made it pretty clear, quite clear in the podcast episode
I’m very sorry if I missed something really. The podcast is 1h30 long can you point to more specific parts addressing this question, and/or give a little sum up to what has been said about it? Thank’s a lot!
That’s not what I think, really…
1 Like
Indeed it may not seem that usefull to re-evaluate more than once per epoch, so re-evalutaion would not be on a daily basis but every 5 days, when epochs are that long.
But as RoS depends on what others do in terms of delegation, you may want to change again depending on the latests delegations that occured and we may even enter an endless loops.
(But it does not change the bot theory anyways)
ok, lets take this example.
lets assume today is day 3, and we’re right in the middle of the epoch. pool x pledge goes up (assuming it makes a positive impact to the ranking, and that other pools don’t do things that would negate it like lowering their costs - factoring in other factors which effect racking which affects rewards) - so ignoring everything else…
then you say ok on day five at the last minute i’ll change and delegate to pool x before day 1 starts. now if one of those pledgers was to retract their pledge - you do realise that (if the pool create blocks within the epoch) that all it takes is that one retraction of that pledge to put at risk the reward that would have gone to the entire pool!
so now that person who thought they were smart is now regretting their decision.
my point here is hypothetical scenarios of gaming the system, not only are they at this point quite low probability, they are also at the mercy of simple human impulse and trust.
for instance the pool operator and those delegating to the pool have to trust that those who have pledged won’t retract their pledge (at least within the epoch) - because everyone loses.
now, i think i have an understanding of why you’re thinking the way you are. you’ve put a lot of money, time and effort into what you now think is going to have been for nothing. i can assure you, you are not the only one.
i can also assure you that most of your concerns would be lessened by watching the video (which has timestamps in the description btw 
). in it they talk about things like the equilibrium points which show that small to medium size pools are the majority (so what works for them is most likely what is going to be the influence the equilibrium that forms A0 and K - having already factored in all the important things like sustainability {healthy rewards to healthy majority} as well as security {sybil attacks etc} ) as well as how, to a person delegating, the difference in which pool to select from a RoS standpoint would be like in the fractions of a difference, it wouldn’t matter so they would most likely end up selecting a pool based on other factors like the values and vision of the pools etc.
3 Likes
now, i think i have an understanding of why you’re thinking the way you are. you’ve put a lot of money, time and effort into what you now think is going to have been for nothing. i can assure you, you are not the only one.
Thank’s a lot for that, and try to understand me! 
Indeed i’m not here to piss people off. Just because I’ve discovered concerns that really worries me about the way Cardano goes, that I did not realize before. And I really hope they are at least going to be answered properly, beginning with why do we not simply limit the ability to receive delegated staked ADA proportionally to pledged ADA, instead of introducing a0.
i can also assure you that most of your concerns would be lessened by watching the video (which has >timestamps in the description btw
I can assure you I did, and carefully, making pauses to keep my concentration up. The whole video is not explaining why this formula was chosen like that, why it is like that, or is better than another, but concentrates on explaining the consequences of the already established formula, and of parameters within it. Anyways I will try to watch it again when I have time.
1 Like
I’m not a stake pool operator, but I do hold ADA and plan to do so for the long term. I’ve subscribed to Hoskinson and numerous other YouTube accounts that put out countless hours of great content and would estimate that I’ve watched over 500 hours of content and read almost every whitepaper I can find. I don’t fully understand everything I’ve read or written even after watching and reading repeatedly, but I do understand I am willing to invest whatever time I feel is necessary to better understand and protect my investment.
Might I respectfully suggest that you watch this 90 minute podcast and read the peer reviewed whitepapers as many times as it takes to first fully understand and then second, be understood when asking your questions with more insight. I’ve personally never seen a project with this much content, transparency, peer review and brainpower at virtually every level inside and outside of the project that is available to get any possible question you have fully answered.
Yes, the information is an incredibly large haystack with a needle of information you may be looking for, but search fields on countless websites and search engines will help you sort through the haystack 
2 Likes
Thank’s for your message, you indeed do well to try and understand as much as possible the crypto you invest in.
Might I respectfully suggest that you watch this 90 minute podcast and read the peer reviewed whitepapers as many times as it takes to first fully understand and then second, be understood when asking your questions with more insight. I’ve personally never seen a project with this much content, transparency, peer review and brainpower at virtually every level inside and outside of the project that is available to get any possible question you have fully answered.
You may indeed do so. But I have a hard time understanding if everything so clear for everyone and many people like you read the whole documentation/papers, and hours of videos, then why why why cannot we have a clear and quick answer to the very simple question. "why do we introduce a0 rather than simply limit the ability to receive delegated staked ADA proportionally to pledged ADA?". If that question can not have a better answer than “it probably is answered somewhere in the doc or videos, read it again” , without pointing to a specific section, then most probably there is a problem.
Anyways I’m at a point were It is more and more possible that I could give up with Cardano, so I’m very sorry but you may understand, at that point, I won’t invest again hundreds or thousand of hours for possibly finding nothing new, sorry. At some point there should be no shifting of the burden of proof. Cardano devs and ambassadors and community should convince people like me their protocol is good, so that I can convince other people later. I’ve told you clearly why I currently doubt it (after having been a supporter of Cardano). If everything is thoroughly thought for years making the best choice after exploring and evaluating all possibilities (as it is supposed to be), then answering my question should be nothing more than a formality or even a copy and past thing.
You are very wrong here. it changes everything. your argument is saying your algo/bot is trying to win a constantly updating lottery based on data from 1-2 epochs ago. literally automation will do nothing for you thats the whole point of VRF’s and the synthetic lottery that underpins Ouroboros. Speculate on about this bot that logically and technically makes 0 sense.
1 Like
That is a good point, and I am not entirely sure of my answer to that but … let’s try :
My best guess is that a proportional limit imposed on the pools will have to consider at each moment the status of each pool regarding delegation. As delegation, re-delegation and/or un-delegation are dynamic in nature, tracking this limit for each pool - when concurrent transactions may or may not delegate to a same pool - seems a daunting and impossible task.
My conclusion is that your solution makes sense on paper, but less so in the implementation and practical realm.
3 Likes
Thank’s a lot for being the first answering the actual question, that’s a real pleasure!
Thing is, as I said in the first post: we do already impose a similar limitation with k, so I really have a hard time understanding why it would be completely ok with k, but impracticable to do it with pledge amount.
What I propose is to replace σ’=min( σ , 1/k ) with σ’= min( σ , 1/k, b0*s ). Really I don’t see where it could be a problem whatsoever.
Goodluck with your superstake algo-bot and rampant speculation on factors you barely understand mate. Cardano will have 1000s of stake pool operators on mainnet making profits soon you can shitpost on the formulas created through peer-reviewed RD or stop whining and join the economic competition 
K is not as such a hard limit, but rather a goal. We could have lower than k pool, even if it makes absolutely no sense. It just incentivizes people to reach k as an optimum, but as we have seen on ITN (k=100), we can have as well much more pools regularly active…
It’s all about a soft-power play. We are always allowed to deviate if we wish to do so.
That is why even with an a0 we don’t like, we can collectively decide not to care too much about it and live our lives happily ever after.
2 Likes
Well it the formula 1/k is an actual solid limit, to the maximum stake you can use to take part of the lottery (it will not prevent you to have more stake, but then having more stake will not give you any more chance to get rewards and blocks). Again the stake taken into account is σ’=min( σ , 1/k ) from the official formula. See also the other post for more insights on formulas.
s delegation, re-delegation and/or un-delegation are dynamic in nature, tracking this limit for each pool - when concurrent transactions may or may not delegate to a same pool - seems a daunting and impossible task.
^Literally this is my entire point. How do you track a synthetic lottery that’s completely dynamic by-design. This algo would not even work if there were no wait period for redelegating to new pools based off this synthetic lottery/dynamic system.
1 Like
As I said, k is fixed and relative delegation can then be calculated once per epoch. With a cap on delegation proportional to pledge, this is hard to do in a dynamic setting. It is easier and safer to shoot on a non-moving target
As I said, k is fixed and relative delegation can then be calculated once per epoch. With a cap on delegation proportional to pledge, this is hard to do in a dynamic setting. It is easier and safer to shoot on a non-moving target
I’m sorry but I really don’t see at all why imposing a variable limit would a problem in anyway. σ’= min( σ , 1/k, b0*s ) would be very easy to compute. And anyway σ’ is only a subfunction of f(σ,s), that already depends on s. You could write my proposal f(σ,s)=R * min( σ , 1/k , b0 * s ), very easy to compute really (clearly not harder than current formula).
And the pledge amount is not supposed to change every other morning either, so degators could have good views on what is the size of the pool before saturation (probably will be much easier for them to predict and understand than knowing when their pool will drop in RoS because of insufficient pledge with current formula).
Anyways the most important question is imagine we do change today the formula as I said in the code, (which I think you will agree is easily doable in just a few lines, If you insist I may even do it in practice on github):
Then on the very practical side what negative thing do you actually expect to happen? I don’t see any.
On the other side if we go with current formula we will get these very negative things:
1-Small pool are disadvantaged over large ones.
2-We encourage everyone to delegate to pools that they don’t know at all but with higher RoS, and to stop delegate to their friend’s or idol’s pool, which increases chances of Sybill attacks.
3- The whole thing is very complicated to understand for everyone, and we can only guess that Incentives will steer pool operators to have sufficient pledge but we cannot be sure of anything, because that’s only incentives and not enforcement.
Which video are you talking about exactly? Just to make sure I didn’t miss it.
Thanks,
Pierre
EDIT: OK , my bad Katsumoto posted it, I need to have my glasses redone haha
2 Likes
Salut Pierre,
un homonyme! ^^
This is this video he refers to:
Pierre.
2 Likes
Merci mon vieux. Ca fait plaisir de retrouver des frenchies ici, je me suis senti très seul les 1ers mois pour organiser des meet-ups et ai finalement laissé tomber devant l’absence totale de participants.
Au plaisir d’échanger de nouveau un de ces 4!
Excellente soirée
1 Like