← Sig Interview Insights

Sig·Software Engineer·Technical Phone Screen·Junior

Junior
May 2026

Summary

SIG quant researcher interview with a probability/expected value question. Pretty classic setup but the edge cases in the calculation can trip you up if you're not careful.

Questions Asked (1)

Q1

You roll a fair 6-sided die twice. If the two rolls are different, you win the higher value in dollars. If they match, you win nothing. What is the expected payout?

Algorithms & Data Structures
Author's notes

Took me a second to set up cleanly.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Use linearity of expectation by defining an indicator for each possible winning value and summing the expected contributions. Alternatively, compute the expected maximum of two distinct rolls by enumerating the 30 ordered pairs where the rolls differ. Clearly state the final expected payout and verify with a quick symmetry or simulation check.

Pro tip: Mention that the expected value can also be derived by considering the expected maximum of two rolls minus the expected value when they are equal, but since equal rolls yield zero, it's simpler to condition on distinct rolls. This shows you understand multiple solution paths and can choose the most efficient one.

1. Define the random variable

Let X be the payout. X = max(R1, R2) if R1 ≠ R2, and X = 0 if R1 = R2.

2. Use linearity of expectation

Express X as the sum over k=1 to 6 of k * I(X = k). Then E[X] = sum_{k=1}^6 k * P(X = k).

3. Compute probabilities P(X = k)

For a given k, X = k occurs when the maximum of the two distinct rolls is k. Count ordered pairs (a,b) with a ≠ b, max(a,b) = k. There are 2*(k-1) such pairs (one roll is k, the other is less than k, and order matters). Total outcomes = 36, so P(X = k) = 2(k-1)/36.

4. Calculate expected value

E[X] = sum_{k=1}^6 k * [2(k-1)/36] = (1/18) * sum_{k=1}^6 k(k-1) = (1/18) * (sum k^2 - sum k) = (1/18)*(91 - 21) = 70/18 = 35/9 ≈ 3.888... dollars.

5. Verify and present

Check by alternative method: condition on distinct rolls. Given distinct, the expected maximum is 14/3? Actually, compute E[max | distinct] = (sum over distinct pairs max)/30 = (sum_{k=1}^6 2k(k-1))/30 = (2*70)/30 = 140/30 = 14/3 ≈ 4.666..., then multiply by P(distinct)=30/36=5/6 to get (5/6)*(14/3)=70/18=35/9. Confirm answer.

Key Points to Mention

  • Linearity of expectation allows breaking down the expected value into sum of probabilities for each outcome.
  • The total number of equally likely outcomes is 36 (6x6).
  • For a specific value k to be the payout, the maximum of the two distinct rolls must be k, which occurs in 2(k-1) ordered pairs.
  • The probability of a tie is 6/36 = 1/6, which contributes zero to the expected value.
  • The final expected payout is 35/9 dollars, approximately $3.89.
  • Alternative approach: condition on the rolls being distinct, compute expected maximum given distinct, then multiply by probability of distinct.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.