Optiver vs Flow Traders vs CTC: OA format comparison
Optiver, Flow Traders, and CTC use different formats, and each needs a different strategy. This page compares their publicly reported length, clock, input type, penalty, and benchmark, and how each one changes the way you sit the paper. Independent comparison; not affiliated with any firm.
The three formats compared
| Optiver-style 80 in 8 | Flow Traders-style | CTC-style OA | |
|---|---|---|---|
| What it is | Pure mental arithmetic | Pure mental arithmetic | Mixed aptitude (four families) |
| Length | 80 questions · 8 min | 75 questions · 8 min | 50 questions · 15 min |
| Seconds / item | 6 | ≈6.4 | ≈18 |
| Input | Typed (classic form) | 4-choice MCQ | Multiple choice |
| Wrong-answer penalty | −1 | −2 | None |
| Skip | Free (0) | Free (0) | Wasted (guess instead) |
| Benchmark provenance | Candidate-reported practice benchmark: ~60 net (not an official cutoff) | TradeMind practice target: 70 net on this 75Q / 8-minute configuration (not a firm cutoff) | No published cutoff |
| Blind-guess EV | N/A (typed) | 0.25×1 + 0.75×(−2) = −1.25 | +0.25 (4-choice), never negative |
Each format has its own scoring system, so practice on one does not fully carry over to the others.
TradeMind web practice scoring reference
This table shows how each simulator on this site calculates its score. It describes our product, not the rule any named firm currently uses. Invitations differ by region, role and recruiting cycle. The benchmark column says whether a number is candidate-reported, a TradeMind training setting, or not publicly available.
| Web practice test | Format | How it is scored | Benchmark source | Worked example |
|---|---|---|---|---|
| Optiver-style 80-in-8 | 80Q · 8 min · Typed | correct − wrong; skips = 0. A doubtful answer can lower net score; skipping is free. | Candidate-reported practice benchmark: ~60 net (not an official cutoff) | 68 correct, 7 wrong, 5 skipped = 61 net |
| Akuna-style 80-in-8 | 80Q · 8 min · Typed | correct − wrong; skips = 0. A doubtful answer can lower net score; skipping is free. | No published pass mark. TradeMind practice target: 60 net (not an Akuna cutoff) | 66 correct, 8 wrong, 6 skipped = 58 net |
| Flow-style practice | 75Q · 8 min · 4-choice MCQ | correct − 2 × wrong; skips = 0. A blind 4-choice guess has −1.25 expected points. | TradeMind practice target: 70 net on this 75Q / 8-minute configuration (not a firm cutoff) | 72 correct, 1 wrong, 2 skipped = 70 net |
| Maven-style practice | 40Q · 6 min · 5-choice MCQ | net = correct − 2 × wrong; benchmark uses correct count. Do not compare the 34-correct benchmark with a net score. | Candidate-reported practice benchmark: ~34 correct on a reported 40Q / 6-minute format (not an official cutoff) | 34 correct, 4 wrong, 2 skipped = 26 net and 34 correct |
| SIG-style practice | 70Q · 8 min · Typed + 4-choice MCQ | correct − typed wrong − 2 × MCQ wrong; skips = 0. The cost of a miss depends on the question's input type. | TradeMind practice target: 55 net on this 70Q / 8-minute configuration (not a firm cutoff) | 60 correct, 2 typed wrong, 1 MCQ wrong = 56 net |
| CTC-style mixed OA | 50Q · 15 min · Multiple choice | raw score = correct; wrong and skipped = 0. With no wrong-answer penalty, a blank cannot beat a guess. | No published cutoff | 38 correct, 10 wrong, 2 skipped = 38 raw |
Optiver-style 80 in 8: typed answers, −1 scoring
The format most candidates mean by "trading math test": 80 arithmetic questions in 8 minutes, with typed answers. Public reports of the classic paper describe −1 per miss, and candidates cite a benchmark of about 60 net (correct minus wrong, skips 0). That is not an official cutoff. 68 correct and 7 wrong is 61 net, not 68.
With no options to pick from, you cannot guess your way to the right number. Points are usually lost to decision latency (half a second spent choosing a method), lapses in negative-marking discipline, and falling accuracy in the last 90 seconds. The rest of Optiver's publicly described OA is a separate set of papers (NumberLogic sequences, Beat the Odds probability, Zap-N games, Zap-Q personality), covered on the 80-in-8 page. 80-in-8 is only one part of the Optiver process.
Some reports describe a 4-choice variant of 80-in-8. The simulator on this site is the typed form. If your invite is MCQ, the arithmetic is the same but the guessing math changes.
Flow Traders-style practice and −2 scoring
Public descriptions of Flow's test do not agree on one configuration; question counts and clocks vary by region, role and recruiting cycle. TradeMind's practice mode uses 75 four-choice questions in 8 minutes, a −2 penalty and a ~70 net training target. Two careless errors cost 4 net points, the harshest scoring in this comparison.
The wrong options are built from common slips: off-by-one, a forgotten carry, adding across a fraction, a decimal point shifted one place. An answer that looks about right is often one of them. The guessing math is also worse than on Optiver-style scoring: a blind 4-choice guess at +1/−2 is worth −1.25 on average, and a guess between the last two options is still worth −0.5. Answer only when you are more than two-thirds sure. Full write-up: Flow Traders math test and how the 70-point practice target works.
CTC OA: mixed reasoning, 15 minutes, no penalty
As candidates describe it, Chicago Trading Company's OA is 50 questions in 15 minutes across numerical reasoning, spatial/abstract, verbal, and logic/attention to detail, with no penalty for wrong answers. That is 18 seconds per item on average. The difficulty is switching between families more than any single question.
The strategy is the opposite of 80-in-8: never leave a blank, and in the last 30 seconds answer everything remaining. Take the fast families first (attention to detail, short syllogisms) and move on from a multi-step depreciation problem. There is no published cutoff; targets discussed online are anecdotal and format-specific. Aim to maximise raw correct answers. Simulator: CTC OA.
What to drill if you do not know which firm yet
- Untimed accuracy on two-digit multiplication, fractions, and decimals. Optiver and Flow both test this. Add a clock only once your error rate is consistently low.
- One full timed 80-in-8 and one full Flow run to feel the difference between typed −1 and MCQ −2. Each page is a fixed set, so your scores are comparable across attempts.
- One CTC-style mixed paper so switching between question types is familiar. Arithmetic drills alone do not train it.
- Skip vs guess as a separate skill. Write down your personal rule before test day. CTC: never leave a blank. Optiver-style (−1): answer when you are more likely right than wrong. This Flow-style mode (−2): answer only when you are more than two-thirds sure, even with two options left.
If you know the firm, focus on its format until you can sit the paper without changing strategy mid-test. The prep guide has a 5-week plan that orders these steps.
Where Zetamac fits
Zetamac is a two-minute arithmetic drill that counts correct operations. It has no skip decision, no penalty and no distractors. A high score shows you can multiply. It does not show you can handle −2 scoring or finish CTC's verbal section. Longer comparison: Zetamac vs firm tests.
Questions
Which of Optiver, Flow Traders, and CTC is the hardest math test?
They test different things, and formats vary. TradeMind's Flow-style practice mode (75 questions, −2 per wrong answer) has the harshest scoring here. Optiver-style 80-in-8 has the tightest clock (6 seconds per question, typed). The CTC-style set tests switching between four reasoning families.
Should I guess?
On CTC, yes: there is no negative marking, so a blank is a wasted chance. On Optiver-style 80-in-8 (−1), skip unless you are confident. On Flow (−2, four choices), a blind guess has an expected value of about −1.25, so skipping matters even more.
Is this official?
No. Formats below are independent reconstructions from publicly described assessments. Firms do not confirm cutoffs. TradeMind is not affiliated with Optiver, Flow Traders, or Chicago Trading Company.
How does TradeMind calculate scores across its web simulators?
The rule depends on the simulator. Optiver- and Akuna-style runs use correct minus wrong. The Flow-style run uses correct minus two times wrong. The SIG-style run subtracts one for a wrong typed answer and two for a wrong MCQ. Maven displays net but evaluates its candidate-reported benchmark by correct count. The CTC-style run is raw correct only. Skips score zero on every listed web simulator.
Can I compare a 60 on one firm simulator with a 60 on another?
No. A score has meaning only with its denominator, clock, input method and penalty. For example, 68 correct and 7 wrong is 61 net on 80-in-8, while 38 correct on the CTC-style set is 38 raw. The worked examples on this page state the unit for each score.
TradeMind is an independent training tool and is not affiliated with, endorsed by, or sponsored by Optiver, Flow Traders, Chicago Trading Company, or any other firm mentioned. Figures are reconstructions from publicly available descriptions of assessment formats and may evolve. Question content on the linked simulators is procedurally generated and does not reproduce actual test material.