Ethereum's built-in randomness for picking who proposes each block, mixed together from validators' signatures.
The idea: Each proposer contributes a value nobody can predict in advance but everyone can verify afterwards: their BLS signature on the epoch number, which is unique for their key. Its hash is mixed into a running total, so the final randomness depends on many proposers. A proposer's only freedom is to skip their block, which gives the last one of an epoch a small say in the outcome.
Why it matters: Fair randomness stops attackers from choosing their own turns.
On the map: SHA-256 mix of BLS reveals. The last proposer can bias 1 bit
Ethereum's randomness for choosing block proposers and committees. Each proposer reveals a BLS signature over the current epoch, and the beacon chain mixes its SHA-256 hash into an accumulated value[1].
The last proposer of an epoch can choose to reveal or skip, so they control one bit of the outcome, a small, well-studied bias. Breaking BLS would let an attacker forge reveals and control the randomness. Verifiable delay functions (VDFs)[2] were proposed to remove the bias and remain a long-term idea.
If any of these breaks, this breaks too.
If this breaks, these break with it.