PRACTICE GUIDE G-RESEARCH
G-Research Part A: what the test is, and how to train for it
The G-Research part a screen is 8 questions in 55 minutes - about 6.9 minutes each - answered by picking one of the options. There is no negative marking, so leaving an item blank gains you nothing over guessing.
Where it sits: Part A of the quant quiz. The practice sitting on this page runs the same item count, the same clock and the same marking rule, with questions generated by Outcry rather than taken from G-Research.
Outcry is not affiliated with G-Research and has no access to their assessment content. This guide describes an assessment format that candidates report publicly; the questions here are generated by Outcry and are not G-Research’s own.
What it screens
A London-based quantitative research and technology company applying data science and machine learning to financial markets.
- ✓Expected value computed fast and used to make a decision
- ✓Conditional probability and Bayes updates without formal notation
- ✓Ranking likelihoods rather than computing them exactly
- ✓Fermi estimation: decomposing an impossible question into three easy ones
Where it sits at G-Research
This section does not arrive on its own. It opens the paper, and the sitting around it runs 90 minutes and 9 items in total, with Part B on the same continuous clock. Candidates who prepare only for this round tend to be caught by the pacing of the rest.
There is no going back. You answer or skip, and the next item loads, which means the decision to leave an item is final at the moment you make it.
The format
These are the numbers the G-Research sitting on this site runs on, matching the format candidates report.
| Questions | 8 |
|---|---|
| Time | 55 minutes |
| Per question | 6.9 minutes |
| Negative marking | No |
| Answer style | Multiple choice |
| Where it sits | Part A of the quant quiz |
What it tests, with a worked example
Every example below is generated by Outcry, drawn from the same question generators the timed drills run. None of them is G-Research’s.
Expected value and odds
Pricing a bet, converting between odds and probabilities, and deciding whether to take it.
Example
You win 20 with probability 65% and lose 20 otherwise. Expected value per bet?
- 6
- 13
- 20
- 0
Answer 6
65% × 20 − 35% × 20 = 6. The edge is the probability gap times the stake.
Conditional probability
Bayes' rule in disguise: a test result, a signal, a drawn card, and what it implies.
Example
Roll a fair die until every face has appeared at least once. What is the expected number of rolls?
Answer 14.7
Coupon collector: 6·(1 + 1/2 + 1/3 + 1/4 + 1/5 + 1/6) = 14.7.
Counting and combinatorics
Arrangements, selections and the complement trick that turns a hard count into an easy one.
Example
A deck has 11 ranks with 5 cards of each rank, 55 cards in all. You are dealt 5. How many different 5-card hands contain exactly four cards of one rank and one card of a different rank?
Answer 2750
Product rule again. Pick the repeated rank 11 ways, pick four of its 5 cards C(5,4) = 5 ways, then the loose card is any of the 50 cards outside that rank. 11 x 5 x 50 = 2,750.
Distributions you should know cold
Binomial, geometric and uniform means and variances, applied rather than quoted.
Example
Two return series correlate at 0.5. What share of the variance in one is explained by a linear fit on the other?
- 100%
- 75%
- 25%
- 70.7%
- 50%
Answer 25%
R² is the square of the correlation: 0.5² = 0.25, so 25%.
Estimation under uncertainty
A number nobody knows exactly, reasoned to the right order of magnitude out loud.
Example
10 people have to get across a river. The boat carries at most 2 and somebody has to row it back each time. Every crossing counts as one trip, in either direction. What is the fewest trips that gets all 10 across?
Answer 17
Count the net progress. A crossing over plus a row back moves 2 across and brings 1 home, a net gain of 1, and costs 2 trips. After 8 such rounds 8 people are across and 2 are left, which the final crossing takes in one go. So 2 x 8 + 1 = 17.
Symmetry arguments
Questions that collapse to one line once you notice the two cases are mirror images.
Example
A fair coin is flipped 4 times. What is the chance every flip is heads, as a percentage?
Answer 6.25
(1/2)^4 = 1/16 = 0.063, or 6.25%.
What a good score looks like
On a paper of 8 questions with no penalty for a wrong answer, the only thing an unanswered question can do is cost you. Probability rounds are usually marked as part of a wider battery, so a standalone threshold is rarely reported. Where candidates do describe one it clusters around getting two thirds of the items right, with the caveat that these sections are short enough that two mistakes move you a long way.
How to train for it
- 01Work the standard results until they are recall, not derivation: complement, conditional, binomial mean and variance, geometric expectation.
- 02Do them timed. A probability question you can solve in four minutes and the paper gives you ninety seconds for is a question you cannot solve.
- 03Sanity-check the magnitude before committing. A probability above one or an expectation outside the range of outcomes is a caught error, and these papers plant both as options.
TRAIN IT HERE
The drills that match each section
Likelihood Ranking
Order events from most to least likely - the exact screen-question format.
Fermi Estimation
Piano tuners in Chicago, atoms in a body - scored on order of magnitude.
Quitters Never Lose
A casino stopping-strategy game that is secretly a martingale lesson.
Tail Risk
Rare events and fat tails, and what they do to a plain expected-value estimate.
SIT THE FULL BATTERY
All the sections back to back on one clock, marked the way the real screen marks them, with a by-skill breakdown at the end. Included with any pass.
Also reported at G-Research
Common questions
- Is the G-Research part a test multiple choice?
- Multiple choice. You pick one option per question.
- How long is the G-Research part a test?
- 8 questions in 55 minutes, which is about 6.9 minutes each.
- Is there negative marking on the G-Research part a test?
- No. A wrong answer costs nothing beyond the mark you would have earned, so leaving an item blank is never better than guessing at it.
- How do I practise for it free?
- Every drill linked on this page is free to play, with no account, inside a daily run cap. Questions are generated fresh each run, so there is nothing to memorise between attempts. The full G-Research Online Quant Quiz sitting puts the sections back to back on one clock.