2

In the paper describing Algorand,two reasons are given to use verifiable random functions/cryptographic Sortion:

Sortition provides two important properties: First, given a random seed, the VRF outputs a pseudo random hash value, which is essentially uniformly distributed between $0$ and $2^{hashlen}-1$. [..] Second, an adversary that does not know $sk_i$ cannot guess how many times user $i$ is chosen, or if $i$ was chosen at all.

Why is the the second property important? What would be the attack scenario if adversaries could predict when user $i$ is chosen?

CryptoThomas
  • 21
  • 1
  • I don't think this subject is off topic because VRF are considered for staking purposes in eth 2.0. – Ismael Jun 29 '19 at 20:34

1 Answers1

1

If validators are known in advance they can be attacked with a denial of service preventing them for fulfilling its role. Delaying the block production or choosing another validator favorable to attackers.

Ismael
  • 30,570
  • 21
  • 53
  • 96