Find the heavy bean in 3 weighings

This classic brain teaser is about using a balance scale in the smartest possible way to find a slightly heavier item among many identical ones. It sits at the intersection of logic puzzles and information theory, which is why it shows up so often in quant prep materials and interview question banks. Top trading firms like questions of this style because they reveal how you structure uncertainty under strict limits.

It trains your ability to maximize information at each step, to reason about discrete outcomes, and to design an optimal sequence of tests under constraints. You practice thinking in terms of search complexity, combinatorics, and trinary branching, all crucial for quantitative problem solving.

This matters in quant interviews because trading and risk decisions often require extracting the most insight from limited data or limited time. Interviewers use this puzzle to see whether you can formalize a seemingly playful setup into an efficient decision process. It's less about arithmetic and more about how you think, a core focus of top quant interviews and serious quant prep.

What it tests

This problem class is governed by the principle of information maximization through partitioning: each weighing on a balance scale can split the set of possibilities into three distinct outcomes (left heavier, right heavier, or balanced), so the optimal strategy is to divide the items into groups that allow each outcome to eliminate as many candidates as possible. The number of weighings required to isolate a unique item among $n$ is determined by how many times you can split $n$ into three roughly equal groups, since each weighing provides up to $\text{log}_3(n)$ bits of information. This is why the minimum number of weighings needed is the smallest integer $k$ such that $n \leq 3^k$. The pattern holds because each weighing is a trinary decision, so the process is analogous to a base-3 search.