How to Beat Nakamoto in the Race

Shu-Jie Cao received a B.E. in Electronic Information Engineering from ShanghaiTech University in 2021. She is currently pursuing a Ph.D. degree at Northwestern University. Her re-search interests include information theory, blockchain, distributed computing, signal processing, and communication networks.

Blockchains like Bitcoin rely on proof-of-work mining, where blocks are created at random times with a memoryless, lottery-like process. The “longest chain” rule then decides which blocks become part of the ledger. But how confident should we be that a block is truly final after waiting for a few confirmations—especially when information does not spread instantly across the network?

In this talk, I will present new results that answer two long-standing questions: (1) what is the most effective strategy an attacker can use to overturn confirmed blocks, and (2) how likely is such an attack to succeed, depending on how long one waits?

Our analysis shows that there is an optimal attack, which we call bait-and-switch, that maximizes an adversary’s chances of overturning confirmed blocks. By carefully modeling the dynamics of block creation and communication delays, we can calculate the exact probability that an attack succeeds for any waiting time.

The results provide a clear, quantitative picture of the trade-offs at the heart of blockchain security: how delays in communication interact with the time users wait for confirmations to determine the true safety of the system.

More details are available in our paper on Shu-Jie Cao and Dongning Guo. 2025. How to Beat Nakamoto in the Race. arXiv:2508.16202, which was also presented at the ACM Conference on Computer and Communications Security (CCS), 2025, Taipei.