Fastest Possible Infection Spread in a Crowd

This brain teaser is about how fast something can spread in an idealized crowd when every interaction is perfectly efficient. It strips away real-world complications to focus on a pure growth mechanism, letting you reason about the best-case dynamics of an infectious process in a random-interaction setting. It's a classic style of question used in quant prep for interviews.

It trains your intuition for exponential growth, branching processes, and upper bounds on stochastic dynamics. You practice thinking about parallel actions, combinatorial explosion, and how to formalize "fastest possible" scenarios. It also sharpens your ability to simplify a noisy probabilistic story into a clean mathematical structure.

This matters for quant interviews because market moves, information flow, and risk cascades often have similar multiplicative behavior. Interviewers want to see if you can quickly identify such structures and reason about scale, timing, and limits under simplified assumptions.

What it tests

This problem class is governed by the principle of exponential growth in branching processes, where each 'active' entity (here, an infected person) independently creates new entities (new infections) in each time step. The key structure is that the number of entities grows multiplicatively, not additively, because every current entity acts in parallel. This doubling pattern arises because each entity's action is independent and can simultaneously produce new entities, leading to $2^n$ growth after $n$ steps. The reason this holds is that the process is unconstrained: there is no overlap or competition for targets, so the maximum possible spread is achieved if each infected person always finds a new healthy person. This exponential pattern is fundamental to many processes in probability, epidemiology, and information spread, whenever each 'agent' can independently create new agents per time unit without interference.