Since proof-of-stake algorithms (like ETH 2) use random sampling to choose the next validator to propose the next block, are there attack vectors with this random selection?

It seems to use RANDAO, a DAO of users returning random numbers. Is this scalable?

1 Answer 1


I am not sure I understand the question, but Eth 2 works this way : during an epoch n, each proposer signs some predetermined verifiable value which will produce a randao reveal (see here). The XOR of every randao reveal of epoch n is the seed of epoch n+2.

Once a seed is known, every validator can know which validator can propose at which slot. So this type of pseudo-randomness is subject to last reveal attack. It means that the last proposer of an epoch can decide whether or not to propose a block (containing her randao reveal) whether the seed favors her or not. An excellent explanation of this is done by Justin Drake here.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.