SIG interview: what candidates actually report
Susquehanna threads are the most distinctive in quant recruiting, because half of them are about poker. That is not a meme: candidates consistently describe a process built around decision quality under uncertainty, with card-game framing that is genuinely load bearing.
This page summarises what candidates have publicly and repeatedly described across forums over recent cycles. It quotes nobody and reproduces no thread; the discussions are linked below so you can read the primary sources yourself. QuantPit is independent and unaffiliated with SIG. Processes change between cycles, roles, and offices, so treat this as a map rather than a schedule, and let your invitation email and the firm's careers page overrule anything here.
What candidates agree on
Reported consistently enough across cycles that you can plan around it.
Questions arrive as bets, not as problems
Candidates repeatedly describe being asked what they would pay, what odds they need, which side they take, and how much they would size. The underlying mathematics is standard probability and expected value; the framing keeps converting beliefs into commitments.
Decision quality is separated from outcome
The most cited theme is being asked about a decision that went badly, or one that went well by luck, and being expected to evaluate the reasoning rather than the result. Candidates who answer with outcomes report it landing poorly.
Poker comes up, but playing is not required
Many candidates describe poker questions or conversations. Those who do not play report doing fine by understanding the underlying concepts: pot odds, break-even frequency, bet sizing, and variance.
The process is often long and multi-round
Accounts commonly describe several rounds with different people, sometimes concentrated into a final day, mixing technical questions with behavioural conversations that keep returning to decision-making.
What candidates disagree on
Where accounts genuinely conflict. Anyone stating these confidently is generalising from one data point, their own.
How much the poker element actually matters
Some candidates describe it as central, others as small talk. Most likely it varies by interviewer, and the concepts are worth knowing either way.
Whether there is an online assessment first
Reports include both an initial test and a direct recruiter conversation, varying by role, region, and application route.
Process length
Accounts range from a couple of weeks to well over a month. Some candidates describe unusually long gaps between stages.
Reported timeline
Commonly several weeks from application to decision, with multiple rounds and sometimes a concentrated final day. Reports of long gaps between stages are frequent enough that silence should not be read as rejection.
Read the discussions yourself
Live searches rather than links to individual threads, so you see current reports rather than a snapshot from an old cycle.
Common questions
Do you need to play poker to get into SIG?
No. Candidates without poker backgrounds are hired regularly. What helps is understanding the concepts poker teaches: pot odds, bet sizing, variance, and information asymmetry.
What is the difference between a good decision and a good outcome?
A good decision chooses the highest expected value with the information available; a good outcome is simply what happened. With large variance the two often disagree, and SIG interviews are widely reported to probe whether candidates can separate them.
How mathematical is the SIG interview?
Mostly undergraduate probability and expected value for trading roles, applied fast and out loud. Research roles carry a higher bar.
What should you prepare for a SIG interview?
Expected value fluency converted into decisions, conditional probability without traps, bet sizing intuition, and two or three honest stories about decisions you made under uncertainty.
Practise the style, free
Reading about it is not preparing for it
The threads tell you what the test looks like. Sitting one tells you whether you would pass it.