Why Randomness is important?

Verifiable randomness is crucial for building trustless blockchain applications like games, lotteries, and fair selection processes. However, generating truly random and verifiable values on-chain is a complex challenge. While the idea of randomness seems simple, its verifiability can add layers of complexity to the application.

The Problem: Many existing solutions for on-chain randomness are vulnerable to manipulation, either by oracles providing the data or by blockchain validators/leaders.

Why can't I just use blockhash?

Using the latest blockhash as a source of randomness is a common, but flawed, approach. Each Solana leader has the power and opportunity to tamper with the randomness. It is trivial for bad-faith parties to collude, making blockhashes inadequate.

What are other solutions?

Ultimately, developers need reliable randomness sources with strong guarantees against manipulation. Let's analyse some common approaches:

Method	Explanation	Adversary(ies)
Oracle Provides Random Value	The Oracle picks a random value for every block.	The Oracle. This defeats the purpose of a decentralised system since it provides total control over the randomness.
Randomness from Elliptic Curve Signatures	A user requests randomness with a unique seed. The oracle responds with Sha256(Ed25519Sign(oracleSigner, seed)).	The Oracle. The Ed25519Signallows a malicious Oracle to produce infinite valid signatures if the user leverages the same nonce, controlling the resulting randomness.
User Secrets (Commit-Reveal Scheme)	User generates a unique secret_key. User commits their wager and Sha256(secret_key). User reveals randomness as Sha256(secret_key + [commit_slot+1].hash)	The User. The scheme creates opportunities for users to collude with block leaders to manipulate the slot hash, undermining the protocol's integrity by leaking the user's key.

Trustless applications are complex, and there is no such thing as a free lunch. Various parties on on-chain applications can exhibit different kinds of adversarial behaviour that ultimately influence randomness. By knowing these common schemes, you can better avoid them.