Single Head Prob Given Past Flips

This classic quant prep question is about updating probabilities when new information rules out some outcomes. You start with a simple random experiment and then learn a partial fact about the result, which forces you to rethink what is still possible. It appears frequently in quant interviews because it looks elementary but exposes whether the candidate really understands conditional probability or is just applying formulas by rote.

It trains conditional probability, sample space reasoning, and rigorous thinking about information. You must be able to translate an English condition into a precise restriction on outcomes and then recompute likelihoods within that reduced universe. It also sharpens intuition about how probabilities can change dramatically when you learn even a single coarse piece of information.

This matters for quant interviews because trading, risk, and statistical modeling all rely on conditioning on partial data. Interviewers use problems like this to probe whether candidates can handle information updates cleanly, avoid common intuitive traps, and reason precisely under uncertainty. Solid performance on such questions signals readiness for more complex Bayesian ideas, filtering logic, and real-time decision-making under noisy signals, all central to quantitative finance roles.

What it tests

When a probability problem introduces new information that restricts the set of possible outcomes (such as 'at least one head'), the correct approach is to condition on this information by redefining the sample space. This means you must exclude all outcomes that contradict the given condition and then recalculate probabilities relative to the reduced set. The underlying structure is that probabilities are always ratios of favorable to possible outcomes, but the definition of 'possible' must reflect all known constraints. This principle holds because probability is fundamentally about relative likelihoods within the set of outcomes that could actually occur, given all information at hand. Failing to adjust the sample space leads to incorrect probabilities because it counts impossible or irrelevant cases.