I was looking into publicly verifiable randomness and noticed that there is a protocol called drand that uses a bunch of nodes that work together to release publicly verifiable randomness at regular intervals.
Why this such a complex protocol required when it is possible to simply use a blockchain's block hash as shown here? This option takes into account miner's being bribed or trying to bias the randomness and seems to be publicly verifiable too.
So why is the second option not good enough?
