Math Questions at Yandex: What to Expect

Math matters at Yandex because the company’s core products—search ranking, recommendation systems, machine learning, and large-scale distributed systems—are built on mathematical foundations. Probability, statistics, combinatorics, and logical reasoning are not abstract concepts here; they directly influence how Yandex designs algorithms to process billions of queries, optimize ad auctions, or train models. The 11 dedicated math questions in their interview process (out of 134 total problems) are a filter for analytical thinking and the ability to translate real-world problems into precise, computable terms. Success here signals you can handle the quantitative rigor required to innovate at scale.

What to Expect — Types of Problems

Yandex’s math questions typically fall into a few key categories. Probability puzzles are common, often involving scenarios with multiple steps or conditional events (e.g., game outcomes, sampling without replacement). Combinatorics problems test your ability to count arrangements or selections under constraints, which is essential for algorithm design. Logical deduction and brainteasers assess structured problem-solving, sometimes without heavy calculation. You may also encounter basic number theory or modular arithmetic concepts applied in coding contexts. The problems are designed to be solved with clear, step-by-step reasoning; brute-force calculation is rarely the goal.

How to Prepare — Study Tips with One Code Example

Focus on core concepts: review probability rules (independence, conditional probability, expectation), counting principles (permutations, combinations), and practice translating word problems into equations. Work on paper first to outline your logic before coding. For many problems, writing a small simulation can verify your reasoning—this is a valuable interview tactic to check your work.

A frequent pattern is using dynamic programming or iterative calculation to solve probability problems. For example, calculating the probability of reaching a point in a grid or the expected number of trials until an event occurs.

def probability_of_reaching(n, p_success):
    # DP array: prob[i] = probability of reaching position i
    prob = [0.0] * (n + 1)
    prob[0] = 1.0  # start at position 0
    for i in range(n):
        if i + 1 <= n:
            prob[i + 1] += prob[i] * p_success
        if i + 2 <= n:
            prob[i + 2] += prob[i] * (1 - p_success)
    return prob[n]

# Example: reach position 5 with step success probability 0.6
print(probability_of_reaching(5, 0.6))

This DP approach efficiently combines probabilities of different paths, a technique applicable to many stochastic process questions.

Recommended Practice Order

Start with foundational probability and combinatorics problems to rebuild intuition. Then, move to Yandex-specific math questions to understand their style. Integrate coding practice for simulation-based verification. Finally, mix math problems with algorithmic challenges, as Yandex interviews often blend these domains. Time yourself to mimic interview conditions.

Practice Math at Yandex