If some group owns 50% of Cardano, what prevents them to from crating two fake pools with 25% and 25% and avoid the penalties? (Or 100 groups with .5%, or whatever is needed.)
(TL;DR: operators can't make more than ~7% annual return on their stake in the long run, and non-operators (people with only 50 ADA for example) can earn ~6-7% by delegating to the most profitable pools.)
Longer answer:
The most you can get from staking is an annual 6-7% return on your stake. I don't fully understand how their saturation parameter works, but here's the general idea: Currently a stake pool with .2% of all ADA staked is considered 'saturated'. This means (in my working understanding, I suspect things are a bit more nuanced) that ADA delegated to them beyond that .2% doesn't receive additional rewards. Therefore it's in the delegators' interest to avoid delegating to pools with greater than .2% of all ADA, and oversaturation works against the stake pool operators as well.
What this means in practice, is that this theoretical group owning 50% of ADA could set up 250 stake pools and fully saturate them for maximum profitability. However, other users can also delegate to those stake pools and share in the rewards. While the operators may receive 1-2% more rewards than non-operating users (there's a limit to how much 'extra' the operators can take, but this extra percentage is intended to incentivize running stake pools), at a certain point of oversaturation (if their pools are profitable enough), it would make more sense for them to just add more stake pools to (or delegate to other pools which are under-saturated) rather than let their existing pools be significantly oversaturated.
Longer answer: The most you can get from staking is an annual 6-7% return on your stake. I don't fully understand how their saturation parameter works, but here's the general idea: Currently a stake pool with .2% of all ADA staked is considered 'saturated'. This means (in my working understanding, I suspect things are a bit more nuanced) that ADA delegated to them beyond that .2% doesn't receive additional rewards. Therefore it's in the delegators' interest to avoid delegating to pools with greater than .2% of all ADA, and oversaturation works against the stake pool operators as well.
What this means in practice, is that this theoretical group owning 50% of ADA could set up 250 stake pools and fully saturate them for maximum profitability. However, other users can also delegate to those stake pools and share in the rewards. While the operators may receive 1-2% more rewards than non-operating users (there's a limit to how much 'extra' the operators can take, but this extra percentage is intended to incentivize running stake pools), at a certain point of oversaturation (if their pools are profitable enough), it would make more sense for them to just add more stake pools to (or delegate to other pools which are under-saturated) rather than let their existing pools be significantly oversaturated.
cardano.org has an insightful blog post: https://iohk.zendesk.com/hc/en-us/articles/900004671183-Chan... and a useful calculator: https://cardano.org/calculator/?calculator=operator
which are both worth looking at if you're interested in knowing more