Every session gets scored on five dimensions: correctness, clarity, speed, rigor, and reflection. Here's what each one actually measures, and why we track them separately instead of collapsing everything into a single number.
Correctness
Did you get the right answer? For calculation problems, that means the right number. For conceptual questions, it means capturing the right idea.
It's the most intuitive dimension and also the least useful to obsess over in isolation. Getting a correct answer through sloppy reasoning is worth less than it looks. Which is exactly what the other four dimensions are for.
Clarity
This one trips people up more than correctness does.
Quant roles put you in rooms with traders, risk managers, and portfolio managers who need to understand complex ideas fast. If you solve the problem but cannot explain why, you are not yet interview-ready. A high clarity score means your explanation was structured, precise, and would make sense to someone hearing the question for the first time.
The most common failure mode: narrating the arithmetic instead of the reasoning. "I multiplied 0.4 by 0.6 and then added" is not clarity. "I'm conditioning on the two exhaustive cases and summing their contributions" is.
Speed
Speed is the most misunderstood dimension. It doesn't reward rushing. It rewards familiarity.
An interviewer has a rough sense of how long a prepared candidate should need to work through a given question. Taking three times as long is a signal. Not that you think slowly, but that you haven't seen enough problems of that type to have internalized the approach.
Real interviews pack 6 to 10 questions into 45 minutes. Pace matters.
Rigor
You can get the right answer for the wrong reason. Rigor is what catches that.
A rigorous answer defines its variables, doesn't skip steps that aren't obvious, handles edge cases, and would hold up if the interviewer asked "why?" at any point. We score it separately from correctness because a shortcut that happened to land is worth less than a sound method that got you there deliberately.
If your correctness is high but your rigor scores are low, you are probably pattern-matching on problem types you've drilled. That works until the framing changes.
Reflection
After you answer, you assess yourself. Reflection scores how accurately you did that.
Interviewers probe this directly. "How do you think that went?" is a real question in real interviews. Candidates who can pinpoint where they stumbled, and articulate what a better answer would have looked like, signal something specific: they know what they know, and they are coachable.
Overconfident self-assessments are just as informative as the answer itself, and not in a good way.
How the Overall Rating Works
It's a weighted average across all five dimensions, adjusted for question difficulty. A strong performance on a hard stochastic calculus question moves your rating more than the same performance on an easy expected value problem.
No single dimension carries you. High correctness with low clarity usually means strong math instincts with a communication problem. That is a specific thing to fix, and fixing it specifically is the whole point of tracking five numbers instead of one.
What to Do With Your Scores
Correctness high, clarity consistently low: stop drilling more problems and start narrating your reasoning aloud until it feels natural. Alone, out loud, every time.
Speed low across a category: that's a volume problem. You need more repetitions of that question type, not slower and more careful thinking.
Rigor dropping in certain areas: that's where your conceptual gaps are hiding. Go back to fundamentals rather than more practice problems.
The number on your dashboard matters less than the pattern underneath it.