← Optiver Interview Insights

Optiver·Data Scientist·Technical Phone Screen·Intermediate

IntermediatePrefer not to say
May 2026

Summary

Quant-style probability round at Optiver for a Data Scientist role. The whole thing was basically one extended brain teaser about waiting for a bus, but they kept layering on complexity and asking you to justify every assumption out loud.

Questions Asked (5)

Q1

A bus arrives exactly every 10 minutes on a fixed schedule. You show up at a random time with no knowledge of the schedule. What is your expected waiting time until the next bus?

Algorithms & Data StructuresTechnical Trade-offs
Author's notes

Seems trivial until you actually have to justify every step.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Recognize that the arrival time is uniformly distributed over the 10-minute interval between buses, so the waiting time is also uniform on [0,10]. Compute the expected value of a uniform random variable on that interval, which is the midpoint, 5 minutes. State the assumptions clearly and note that the result is independent of the schedule's starting point.

Pro tip: Explicitly state that you assume a fixed schedule with no randomness in bus arrivals, and that your arrival time is uniformly distributed. This shows you understand the model and avoids ambiguity.

1. Define the random variable

Let W be the waiting time until the next bus. Since you arrive at a random time with no knowledge of the schedule, W is the time from your arrival to the next bus.

2. Identify the distribution

Because buses arrive every 10 minutes, the time until the next bus is uniformly distributed between 0 and 10 minutes. This assumes your arrival time is uniformly random over the interval.

3. Compute the expected value

For a uniform distribution on [0,10], the expected value is the midpoint: (0+10)/2 = 5 minutes.

4. State the final answer

The expected waiting time is 5 minutes. Emphasize that this is a theoretical average and actual waiting times will vary.

Key Points to Mention

  • Uniform distribution of arrival time over the 10-minute interval
  • Expected value of a continuous uniform distribution is the midpoint
  • Assumption of a fixed, deterministic schedule with no randomness in bus arrivals
  • Independence of the result from the schedule's phase (e.g., whether buses arrive at :00, :10, etc.)
  • The answer is 5 minutes, but actual waiting times can range from 0 to 10 minutes
  • Potential real-world deviations: bus delays, variability, or non-uniform arrival patterns

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

Q2

Now each bus independently has a 50% chance of running late. The definition of 'late' is left vague. State a precise model, justify it, compute the expected wait, and then explain how the answer changes under a different reasonable model.

Algorithms & Data StructuresAdaptability & AmbiguityTechnical Trade-offs
Author's notes

This is where I got tripped up.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by explicitly defining a precise model for bus lateness, such as each bus independently being late by an exponential amount with probability 0.5, and justify why this is reasonable. Compute the expected waiting time under this model, then discuss how the answer changes if you assume a different distribution, like a uniform delay, highlighting the impact of distributional assumptions.

Pro tip: Always state your assumptions clearly and connect them to real-world scenarios; interviewers value candidates who can reason under ambiguity and adapt their models when new information is presented.

1. Define a precise model

Choose a specific probability distribution for the delay when a bus is late (e.g., exponential with rate λ) and specify that each bus is independently late with probability 0.5. Justify the choice by relating it to typical bus arrival patterns or simplicity.

2. Compute expected wait

Derive the expected waiting time for a passenger arriving at a random time, considering that the bus may be on time (zero delay) or late (positive delay). Use the law of total expectation.

3. Explain alternative model

Introduce a different reasonable model, such as a uniform delay between 0 and 10 minutes when late, and compute or qualitatively describe how the expected wait changes. Highlight that the expected wait depends on the distribution's mean and variance.

4. Discuss implications and trade-offs

Compare the models, noting how assumptions affect the result and which model might be more realistic. Mention that in practice, data would be used to estimate parameters.

Key Points to Mention

  • Independence of buses and the 50% late probability.
  • Choice of delay distribution (e.g., exponential, uniform) and its parameters.
  • Law of total expectation to compute overall expected wait.
  • Sensitivity of expected wait to distributional assumptions (mean and variance).
  • Real-world relevance: bus delays are often not memoryless, so exponential may be unrealistic.
  • Potential use of historical data to calibrate the model.

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

Q3

If buses arrived as a Poisson process with the same mean headway of 10 minutes, what would the expected wait be, and why does memorylessness change the answer compared to the fixed-schedule case?

Algorithms & Data StructuresTechnical Trade-offs
Author's notes

Follow-up that came fast after Part 2.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the two scenarios: fixed 10-minute headway means a uniform wait between 0 and 10 minutes, averaging 5 minutes. For a Poisson process with mean headway 10 minutes, the arrival rate is 0.1 per minute, and by memorylessness the wait time is exponential with mean 1/λ = 10 minutes. Explain that memorylessness makes the expected wait equal to the mean interarrival time, not half of it, because the residual time distribution is the same as the full interarrival distribution.

Pro tip: Emphasize that the Poisson process is a model for random arrivals, and in practice bus arrivals are not perfectly Poisson due to scheduling; but for the interview, show you understand the mathematical distinction and can connect it to real-world implications like passenger wait times.

1. Define the fixed-schedule case

State that with a fixed 10-minute headway, wait times are uniformly distributed between 0 and 10 minutes, so the expected wait is 5 minutes.

2. Define the Poisson process case

Specify that a Poisson process with mean headway 10 minutes has rate λ = 1/10 per minute, and interarrival times are exponential with mean 10 minutes.

3. Apply memorylessness

Explain that due to the memoryless property of the exponential distribution, the expected residual wait time given any arrival time is still 10 minutes, not 5.

4. Compare and contrast

Contrast the two results: fixed schedule gives 5 minutes, Poisson gives 10 minutes, highlighting that variability increases average wait.

5. Discuss implications

Mention that this is a classic result (e.g., inspection paradox) and that real-world bus systems often have less variability than Poisson, so waits may be between 5 and 10 minutes.

Key Points to Mention

  • Fixed schedule: uniform wait distribution, expected wait = half of headway = 5 minutes.
  • Poisson process: exponential interarrival times with mean 10 minutes, rate λ = 0.1 per minute.
  • Memorylessness: the expected residual wait time is equal to the mean interarrival time (10 minutes).
  • The inspection paradox: a randomly arriving passenger is more likely to land in a longer interval, increasing average wait.
  • Real-world bus arrivals are not perfectly Poisson; they often have some regularity, so actual waits may be between 5 and 10 minutes.
  • The result generalizes: for any renewal process, the expected waiting time is E[T^2]/(2E[T]), which for exponential gives E[T].

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

Q4

In the fixed-delay model where a late bus is delayed by exactly delta minutes, how does the expected wait behave as a function of delta? What happens when delta equals 10?

Algorithms & Data StructuresTechnical Trade-offs
Author's notes

Didn't love this one in the moment.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

First, clarify the fixed-delay model: a bus that is late is delayed by exactly δ minutes, while on-time buses are not delayed. Then derive the expected wait as a function of δ, showing that it increases linearly with δ. Finally, evaluate at δ = 10 to give a concrete answer, noting any assumptions about bus arrival distribution.

Pro tip: Mention that the result depends on the probability of a bus being late; if that probability is p, the expected wait increases by p·δ. This shows you understand the underlying stochastic model and can generalize.

1. Define the model

State the assumptions: buses arrive according to a known schedule, but with a fixed delay δ applied only to late buses. Clarify the probability p that a bus is late.

2. Derive expected wait

Express the expected wait as a function of δ, typically as a linear function: E[wait] = base_wait + p·δ, where base_wait is the expected wait without delays.

3. Analyze behavior as δ varies

Discuss how the expected wait changes with δ: it increases linearly, with slope equal to the probability of a late bus. If p=1, the slope is 1; if p<1, the slope is less than 1.

4. Evaluate at δ = 10

Plug in δ=10 to compute the expected wait. If p is known, give a numerical answer; otherwise, express it in terms of p.

5. Discuss implications and edge cases

Mention that if δ is very large, the expected wait grows without bound. Also note that if the bus is never late (p=0), δ has no effect.

Key Points to Mention

  • The expected wait increases linearly with δ, with slope equal to the probability of a late bus.
  • If the probability of a late bus is p, then E[wait] = E[wait without delays] + p·δ.
  • At δ=10, the expected wait increases by 10p minutes compared to the no-delay scenario.
  • The result assumes that delays are independent and that the bus schedule is otherwise regular.
  • If p=1 (all buses late), the expected wait increases by exactly δ minutes.
  • The analysis highlights the importance of punctuality: reducing p or δ directly reduces average waiting time.

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

Q5

Argue that among all stationary arrival processes with a mean gap of 10 minutes, the perfectly regular schedule minimizes a random passenger's expected wait.

Algorithms & Data StructuresTechnical Trade-offs
Author's notes

They saved this for the end and I was already a bit fried.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Use the inspection paradox and renewal theory to show that expected wait depends on the second moment of the interarrival distribution. Since the mean is fixed at 10, minimizing the second moment (and thus variance) minimizes the expected wait. The deterministic (perfectly regular) schedule has zero variance, so it achieves the minimum.

Pro tip: Connect the result to real-world systems: this explains why scheduled buses with low variance reduce average wait, and why high-variance arrivals (e.g., Poisson) increase it. Mentioning the inspection paradox shows depth.

1. Define the model and objective

State that we consider a stationary arrival process with mean interarrival time E[X] = 10 minutes. A random passenger arrives at an arbitrary time and waits until the next arrival; we want to minimize E[W].

2. Apply the inspection paradox

Explain that a random passenger is more likely to land in a longer interarrival interval. The expected length of the interval containing the passenger is E[X^2]/E[X], not E[X].

3. Derive expected wait

For a stationary renewal process, the expected waiting time is E[W] = E[X^2]/(2E[X]). Since E[X] = 10 is fixed, minimizing E[W] is equivalent to minimizing E[X^2].

4. Minimize the second moment

By Jensen's inequality, E[X^2] ≥ (E[X])^2, with equality iff X is constant. Thus the minimum occurs when X = 10 deterministically, i.e., a perfectly regular schedule.

5. Conclude and interpret

Therefore, among all stationary arrival processes with mean gap 10, the deterministic schedule minimizes the random passenger's expected wait. This is a classic result in renewal theory.

Key Points to Mention

  • Inspection paradox: random arrival is biased toward longer intervals.
  • Renewal theory formula: E[W] = E[X^2]/(2E[X]).
  • Jensen's inequality: E[X^2] ≥ (E[X])^2, equality iff X is constant.
  • Deterministic schedule has zero variance, hence minimal second moment.
  • Comparison to Poisson process: exponential interarrival times give E[W] = E[X] = 10, which is larger than the deterministic case (5 minutes).
  • Stationarity assumption ensures the process is in equilibrium.

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