← SIG (Susquehanna) Interview Insights

SIG (Susquehanna)·Software Engineer·Technical Phone Screen·Junior

Junior
Jun 2026

Summary

SIG quant engineer interview, probability question that looks straightforward but bites you if you're sloppy with the base rates.

Questions Asked (1)

Q1

30% of posts are fake news. Fake posts have an 80% chance of getting 100+ likes, while real posts have an 8% chance. A randomly selected post has 100+ likes. What's the probability it's fake news?

Algorithms & Data StructuresProduct Analytics & Metrics
Author's notes

Classic Bayes setup but the numbers trip you up if you don't slow down.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Recognize this as a Bayes' theorem problem and clearly define the events and probabilities. Compute the posterior probability using the formula P(Fake | 100+ likes) = P(100+ likes | Fake) * P(Fake) / P(100+ likes), where P(100+ likes) is the sum of the probabilities of having 100+ likes given fake and given real, weighted by their priors. Finally, simplify the fraction to get the answer.

Pro tip: Always state your assumptions (e.g., that the 30% base rate applies to the randomly selected post) and double-check the arithmetic. In interviews, explaining the reasoning step-by-step is often more important than just giving the final number.

1. Define events and probabilities

Let F be the event that a post is fake news, and L be the event that a post has 100+ likes. Write down the given probabilities: P(F) = 0.30, P(L|F) = 0.80, P(L|F^c) = 0.08.

2. Compute total probability of L

Use the law of total probability: P(L) = P(L|F)P(F) + P(L|F^c)P(F^c) = 0.80*0.30 + 0.08*0.70 = 0.24 + 0.056 = 0.296.

3. Apply Bayes' theorem

Compute the posterior probability: P(F|L) = P(L|F)P(F) / P(L) = 0.24 / 0.296.

4. Simplify and interpret

Simplify the fraction: 0.24/0.296 = 240/296 = 30/37 ≈ 0.8108. So the probability is about 81.1%. Interpret the result in context.

Key Points to Mention

  • Bayes' theorem and its application to conditional probability
  • Law of total probability to compute the denominator
  • Correctly identifying the base rate (prior probability) of fake news
  • Avoiding the base rate fallacy by not ignoring the prior
  • Simplifying the fraction to get an exact answer (30/37) and then converting to a percentage
  • Checking that the result makes intuitive sense (since fake posts are much more likely to get 100+ likes, the posterior should be high)

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