Is $10m Ada the magic number?

From the delegators perspective, there is one single metric that makes one pool more profitable than another, which is “cost per block” - with luck and runtime stability being all equal. Cost per block is the share of block rewards that goes to the pool rather than to the delegators - this is explored here.

When we talk about cost to the operator vs. cost to the delegator (i.e. is it more profitable to run my own pool). There is a break even point, which can be well below 10m depending on cost for the infrastructure and how you value your skill/time - this is explored here.

Recently another metric came into the game, which is “effective cost per block”. This is when an operator realizes that the pool reward (and hence the cost to the delegator) is much higher than is actually needed to run the pool. In that case an operator may do extra payouts to delegators to lower the cost per block, so that even a small pool can stay competitive. Various such schemes exist, which to the extreme would put a delegator on exact equal ROA to the operator. Running a pool is of course not free and that would need to be covered in any case.

Lastly, we are talking about differences between 4-6% ROA. Yes, a 30% cut on reward is significant, but in the overall scheme of things even 6% ROA will be totally dwarfed by price movements of the underlying asset.

You may need to shop around a little. There is often good reason to choose a pool not just on ROA. For example, whether it does other good things in the world - does it “Make the world work better for all”? Some good options are here.