“Algorand: Scaling Byzantine Agreements for Cryptocurrencies” is a paper by Jing Chen and Silvio Micali (MIT), first circulated in 2017 and formally published at ACM SOSP 2019 (the premier systems conference). Micali, a Turing Award winner and co-inventor of zero-knowledge proofs, designed Algorand to solve a problem he called the “trilemma” before Buterin popularized that term: achieving security, scalability, and decentralization simultaneously in a public blockchain.

The protocol’s central innovation is cryptographic sortition — a technique where each node secretly and privately determines whether it has been selected as a committee member for the current round, using a Verifiable Random Function (VRF), revealing that membership only when submitting its vote, making targeted denial-of-service attacks impossible.

> PDF hosting: The theoretical overview is hosted at the Algorand Foundation’s CDN and on the official algorand.com. The full technical paper is available via ACM Digital Library and arXiv. All versions are freely distributable.

---

Publication and Context

In 2017, every major blockchain faced a dilemma: Bitcoin and Ethereum used proof-of-work, which was secure but slow and energy-wasteful. BFT-based alternatives (Tendermint, HotStuff) were fast but required small, known validator sets. Nobody had produced a permissionless system that was both fast and provably secure with a large validator set.

Micali had spent years working on the problem before founding Algorand Inc. in 2017 and launching the mainnet in 2019. The paper’s academic rigor — backed by formal proofs rather than engineering heuristics — distinguished it from contemporaries.

Key facts:

• First circulated: May 2017
• ACM SOSP publication: October 2019
• Mainnet launch: June 2019
• Block time: ~3.3 seconds (practical; faster than any PoW chain)
• Finality: Deterministic within one round (~3–5 seconds)

---

The Core Innovation: Cryptographic Sortition

Traditional BFT consensus selects a fixed committee in advance — making those committee members known targets. If an attacker can identify and corrupt or DoS members before they act, the consensus breaks.

Algorand’s solution: no one knows who is on the committee until committee members announce themselves.

How sortition works:

1. Each user holds a secret key and the current blockchain seed (public randomness)
2. Each user evaluates VRF(sk, round || role || seed) — a Verifiable Random Function
3. The VRF output is a pseudorandom value that determines: did I get selected? How many votes do I have?
4. If selected, the user publishes their VRF output + proof alongside their block proposal or vote
5. Anyone can verify the proof (VRF is publicly verifiable) but no one can predict the output in advance

Because selection is revealed only during participation, there is no “committee” to target before the round begins. By the time an attacker learns who was selected, the round is already over.

---

BA★: Binary Byzantine Agreement

Once block proposals arrive, agreement is reached via BA★ — a simplified BFT protocol optimized for Algorand’s assumptions:

1. Block proposal phase: Sortition selects ~20 proposers. The lowest-value VRF output wins (serves as a priority). Nodes gossip the proposed block.
2. Soft vote phase: A committee of ~2,000 users votes to “soft-commit” to the block with the best priority.
3. Certify vote phase: A second committee votes to certify the block. If 2/3 agree, the block is committed.
4. Fallback (GC rounds): If no 2/3 supermajority is reached (due to network partition or timing), the protocol runs recovery rounds using binary BA.

Key property: BA★ achieves agreement in expected O(1) rounds — usually 2 phases. There are no forks; once a block is committed, it cannot be reversed.

---

Sections of the Whitepaper

Section	Content
1. Overview	The trilemma; why PoW and traditional BFT fail
2. System Model	Network model; honest majority assumption
3. Cryptographic Sortition	VRF construction; weighted stake selection
4. BA★	Binary Byzantine Agreement protocol; the two-step process
5. Block Proposal	How proposers are selected and blocks gossipped
6. Agreement	Soft vote → certify vote → commit
7. Properties	Safety and liveness proofs under the honest majority assumption
8. Performance	Theoretical throughput analysis

---

Security Assumptions

Assumption	Detail
Honest majority	>2/3 of staked ALGO is held by honest users
Network synchrony	Messages are delivered within a known time bound (partial synchrony)
Honest majority at round	Must hold at the moment of committee selection

The 2/3 honest majority requirement is weaker than Bitcoin’s 50%+1 hash power requirement — but it’s a stake-based majority, not a node-count majority. A single whale holding 33%+ of ALGO could theoretically break safety.

---

Reality Check

The formal proofs are solid. Algorand’s fork-free consensus is genuinely achieved in production — the mainnet has run since 2019 without a fork. Block times of ~3.3 seconds with deterministic finality put it ahead of Ethereum (historically) and Bitcoin in finality speed.

However, Algorand faced adoption challenges despite technical excellence. The ALGO token’s distribution (heavy VC/foundation allocation) and early inflation schedule drove token price down over 2021–2023. The perception that Algorand was “technically perfect but commercially weak” became a recurring critique on Crypto Twitter.

AVM (Algorand Virtual Machine) supports smart contracts and ASAs (Algorand Standard Assets) but has seen less developer adoption than EVM-compatible chains.

---

Legacy

Micali’s VRF-based sortition influenced Ethereum’s validator selection design and subsequent PoS research broadly. The concept of “ephemeral committees unknown in advance” became a key design principle for randomness in consensus. Algorand’s consensus is used as a benchmark in academic papers comparing PoS designs.

---

Related Terms

• Algorand
• Stellar Whitepaper
• Cardano Whitepaper
• Avalanche Whitepaper
• Ethereum Beacon Chain Whitepaper

---

Research

• Chen, J., & Micali, S. (2019). Algorand: Scaling Byzantine Agreements for Cryptocurrencies. SOSP 2019.

— Primary source. The ACM published version is the definitive reference.

• Gilad, Y., Hemo, R., Micali, S., Vlachos, G., & Zeldovich, N. (2017). Algorand: Scaling Byzantine Agreements for Cryptocurrencies. arXiv:1607.01341.

— Earlier arXiv version; includes additional appendices on the protocol.

• Micali, S. (2016). ALGORAND: The Efficient and Democratic Ledger. arXiv:1607.01341.

— The original 2016 circulated draft with the core VRF sortition concept.