Symmetry Arguments in quant interviews
The problems that collapse in one line — if you notice the symmetry.
TRY A PROBLEM NOW · NO SIGNUP
In the classical secretary problem, as n → ∞, what fraction should you reject outright before accepting the next best-so-far candidate?
WHAT IT IS
Symmetry arguments exploit structural equivalences in a problem to avoid direct computation. The classic form: if positions, players, or outcomes are exchangeable, their probabilities or expectations must be equal, and together they must sum to some known total. In quant interviews, symmetry questions are the ones that distinguish candidates with strong probabilistic intuition from candidates who reach for brute force. A well-posed symmetry question looks like it requires pages of casework; a candidate who sees the symmetry solves it in one line. Interviewers test symmetry partly as a proxy for probabilistic maturity — it reveals whether the candidate thinks structurally about problems or just computes mechanically. Common symmetry patterns include: exchangeability of players in a game, rotational symmetry in circular arrangements, reversibility in random walks, and equivalence of events under relabelling.
WHEN IT APPEARS IN INTERVIEWS
Most common at Jane Street, where symmetry-based probability questions are a signature style. Also shows up at SIG, Citadel, and Optiver, usually in later rounds where interviewers test depth rather than speed.
FIRMS THAT TEST THIS
SAMPLE PROBLEMS
n people sit in a circle. Each is given a random hat, red or blue, each with probability 1/2 independently. What is the expected number of adjacent pairs with different colours?
By symmetry, any adjacent pair has P(different colours) = 1/2. There are n adjacent pairs. E[different-coloured pairs] = n · 1/2 = n/2. The apparent complexity (correlations between adjacent pairs) is irrelevant — linearity plus symmetry gives the answer in one line.
You deal 5 cards from a standard deck to yourself, and then 5 cards to an opponent. What's the probability your hand is better than your opponent's?
By symmetry, P(you win) = P(opponent wins). P(tie) is roughly negligible. So P(you win) ≈ 1/2. You don't need to enumerate poker hand rankings.
A symmetric random walk starts at 0 on the integers. What's the probability it ever returns to 0?
By symmetry and the law of large numbers, the walker crosses every integer infinitely often (recurrence of 1D symmetric random walk). So P(return) = 1. The symmetry argument: the probability of being 'stuck' above 0 is zero because the walker can always be pulled back by symmetry of the step distribution.
You draw three independent uniform(0,1) random variables X₁, X₂, X₃. What's the probability they come out in increasing order, X₁ < X₂ < X₃?
There are 3! = 6 equally likely orderings of the three values (by exchangeability). Only one of them is 'strictly increasing'. P = 1/6. No integration required.
SOLVING STRATEGIES
- ·Ask: are there permutations of players, positions, or outcomes that leave the problem unchanged? If yes, their probabilities are equal.
- ·Use symmetry + total probability to extract individual probabilities. If n symmetric events partition the sample space, each has probability 1/n.
- ·Check reversibility. A symmetric random walk looks the same forwards and backwards; this breaks many seemingly hard questions open.
- ·When stuck, try to make the problem 'more symmetric' via clever relabelling — often the hardest half of the trick is seeing the symmetry at all.
- ·Always verify symmetry rigorously before using it. Interviewers will push back if the symmetry claim is fuzzy.
COMMON VARIATIONS
- ·Exchangeability: sequences of random variables that are invariant under permutation.
- ·Reversibility: Markov chains that look the same time-reversed.
- ·Symmetry in game theory: minimax strategies often emerge from symmetry.
- ·Dihedral / rotational symmetries in geometric probability problems.
FAQ
Before computing, spend 30 seconds asking 'what would change if I swapped X and Y?' or 'what would change if I relabelled players?'. If the answer is 'nothing', you likely have symmetry.
Not always — but it often reduces the problem to a form where clean answers are visible. When it doesn't give the whole answer, it usually cuts the problem in half.
They pair constantly. Linearity decomposes a random variable into a sum of indicators; symmetry makes many of those indicators have equal probability, which collapses the sum to a single multiplication.
They test whether candidates think about problems structurally. Symmetry is hard to fake — either you see it or you don't — so it's a clean filter for probabilistic maturity.
RELATED TECHNIQUES
The single most common technique in quant interviews. Every problem reduces to it eventually.
"Given X happened, what's the probability of Y?" — the second-most-common framing in quant interviews.
E[X + Y] = E[X] + E[Y], independent or not. The most over-powered tool in probability.
Not ready to sign up? Get 5 hand-picked quant interview problems every week. No spam, unsubscribe any time.
Drill symmetry arguments problems now
400+ problems, adaptive selection, AI-generated alternative explanations. Free tier covers 20 attempts.
Start practicing free