Hermann Ebbinghaus documented this in 1885, and nothing has changed: without review, most of what you study in a session is gone within 24 hours. Nearly all of it within a week.
This matters for quant prep because the material spans hundreds of problem types across seven categories. You cannot hold all of them in active memory at once, and you cannot hold any single one for long without deliberate review. The question is how you structure that review.
What Spaced Repetition Actually Does
The core idea is simple. Reviewing material at increasing intervals is far more efficient than reviewing it repeatedly in a short window.
When you successfully recall something, the memory trace strengthens and the next review can wait longer. Miss a review, the interval shrinks. Over time, material you have mastered needs only occasional check-ins. Stuff you keep getting wrong stays in rotation more frequently.
The practical effect is a self-correcting schedule that prioritizes your weak spots automatically. You show up, it tells you what needs attention.
Not All Quant Material Forgets at the Same Rate
There are three layers of knowledge in quant prep, and they decay at different speeds.
Conceptual frameworks are the stickiest. Bayes' theorem setup, the martingale property, put-call parity. Once genuinely understood, these hold up well between reviews.
Problem archetypes sit in the middle. The secretary problem, the birthday problem, random walk variations. You need to recognize these on sight, and they need moderate maintenance to stay sharp.
Calculation procedures are the most fragile. Ito's lemma mechanics, moment-generating function algebra, conditional expectation steps. These fade quickly and need the most frequent review.
A good study schedule treats these three layers differently rather than cycling through everything at the same rate.
Why Cramming Feels Like It Works
The illusion has a name: massed practice. After three hours on expected value problems, you can do them fluently. That fluency is real. It is also mostly gone by next week.
The more insidious problem is that cramming builds confidence on the canonical version of a problem. Interviewers know this. They reframe. They change the setup. They follow up in ways that probe whether you understand the mechanism or just learned the steps.
Spaced repetition builds different fluency. You have seen the concept in enough contexts, with enough time between exposures, that it generalizes when the framing changes.
A Practical Schedule
No software required. A realistic structure for one category:
- Day 1: Study the category in depth
- Day 3: Review the hardest third of what you covered
- Day 7: Light review of the full category
- Day 14: Spot-check with a handful of problems
- Monthly: Full pass across all categories
One thing people get wrong: the discomfort of letting material sit unreviewed is not a sign the system is failing. It is the mechanism. The retrieval effort on day 7 is what consolidates the memory. Day 1 just introduces it.
Interleaving
Mixed practice beats blocked practice. After doing 20 conditional probability problems in a row, everything starts to look like a conditional probability problem. In an interview, there is no such context. You have to identify what you are dealing with before you can solve it.
Interleaving different problem types within a session trains exactly that recognition. It is slower and more frustrating than blocked practice. It is also what actually prepares you for the real thing.
Spacing over time plus interleaving within sessions. That combination produces knowledge that holds up under pressure, not just knowledge that holds up until Thursday.