How to learn with this course
Learn the loop that each lesson uses and why a self-test beats a second read. Then take a 10-question pretest to see where you start.
In this lesson you will learn
- The parts of each lesson, and why each part is there
- Why a self-test helps your memory more than a second read
- How to use the worksheets, the glossary, and the review deck
- What you know about startup funding today, from a 10-question pretest
Most founders learn startup funding at the worst possible time: the week before they sign. This course lets you learn it earlier. You make small mistakes in problems, not large mistakes in contracts. This first lesson shows how the course works, and why it asks you to do more than you read.
Start with a question about memory. Which plan gives better memory one week later: study a text in four sessions, or study it in one session and test yourself three times?
Predict first
In a 2006 study, students who read a short text approximately 14 times remembered 40% of it one week later. A second group read it approximately 3.4 times and then did three recall tests. What percentage did the second group remember?
Type a guess before you continue. A guess first, even a wrong one, makes the answer easier to remember.
The test group remembered 61%, against 40% for the group that only read. The group that only read also expected to remember more. A second read feels like learning, but a test does more for your memory (Roediger and Karpicke, 2006).
Why a self-test works better than a second read
When you test yourself, you must find the idea in your memory without the text in front of you. That effort is the practice, and its name is retrieval practice. A second read gives you the idea again, so you do not have to find it.
Dunlosky and his colleagues (2013) rated ten study methods. Only two got a “high utility” rating: practice tests and practice that you spread over time. A second read, highlights, and summaries got a low rating. In a review of 225 studies, Freeman and his colleagues (2014) compared lecture courses with courses that use active learning. The failure rate was 33.8% in lecture courses and 21.8% in courses with active learning.
The loop in each lesson
Each lesson uses the same loop, so that you always know what comes next. The table shows each part, what you do, and why the part is there:
| Part | What you do | Why it helps |
|---|---|---|
| A question | Read a question or a headline that you want to understand | A question gives the lesson a purpose. |
| Predict first | Type a guess before you see the answer | A committed guess makes the answer easier to remember, also when the guess is wrong. |
| Short explanations | Read one idea at a time, each with a numeric example | A number lets you do a check of each idea. |
| Worksheets | Move the numbers in a widget and read what they mean | You see how the full system responds, not one example only. |
| A worked example | Show one step at a time, and try each step in your head first | Beginners learn more from solved examples than from problems without help. |
| Work it out | Do the last steps of a problem yourself | The steps that you do yourself move you from examples to full problems. |
| Quiz | Answer 4 to 6 questions and read the feedback | A test with feedback stays in memory longer than a second read. |
Several parts come from research results. Hohman and his colleagues (2020) describe a study in which a prediction that learners drew made their recall better. Nicky Case calls this pattern “Place Your Bets.” Renkl and his colleagues (2002) studied worked examples. They removed the solved steps one at a time, and learners did better on similar problems than after a sudden change to full problems.
Why the feedback names your error
Each wrong option in a quiz is a mistake that founders make. When you select it, the feedback tells you which error you made, for example: “This divides by the pre-money valuation.” The feedback does not only say “Try again.”
This choice is important. In a review of 607 effects, Kluger and DeNisi (1996) found that more than one third of the feedback in the studies made performance worse. Feedback about the task helps more than praise (Hattie and Timperley, 2007). Feedback after a multiple-choice test also decreases the risk that you later remember a wrong option as correct (Butler and Roediger, 2008).
Spread your reviews over time
Your chance to remember a new idea decreases fast at first, and then more slowly. One model of this decrease comes from Settles and Meeder (2016), who used it to schedule reviews at Duolingo. In the model, your chance to remember an idea halves with each “half-life” that passes. Move the numbers to see how fast the chance decreases.
Model
How fast you forget a new idea
In the half-life model, your chance to remember an idea halves with each half-life that passes. Change the half-life and see when a review is due.
Days until the chance to remember falls to 50%.
- Chance after 1 day
- 84%
- Chance after 1 week
- 30%
- Review is due after
- 1.3 days
With a half-life of 4 days, you remember a new idea after one day with a chance of 84%, and after one week with a chance of 30%. To keep the chance at 80% or more, review after 1.3 days.
The review deck of this course also spreads reviews over time. A card that you know waits longer each time before it comes back: 1 day, then 3, then 7, then 21 days.
The best time for a review depends on how long you want to remember. Cepeda and his colleagues (2008) tested one review before a final test:
- For a test one week later, the best gap was approximately 20% to 40% of that week: 1.4 to 2.8 days.
- For a test one year later, the best gap was approximately 5% to 10% of the year: 18 to 37 days.
Do the math for two cases yourself.
Work it out
An idea has a half-life of 4 days. What is your chance to remember it 8 days after you last practiced it?
Chance = 2^(−8 ÷ 4) = 2^(−2) = 25%. After one half-life (4 days), the chance is 50%. After the second half-life, it is half of 50%.
Work it out
You want to remember an idea for one year. The best first review comes after 5% to 10% of that time. What is the earliest of those days, in days after the lesson?
365 × 0.05 = 18.25 days, so approximately 18 days. The latest good day is 365 × 0.10 = 36.5 days, approximately 37 days.
The review deck of this course uses spaced repetition. When you complete a lesson, its terms go into your deck. A card that you know waits longer before it comes back: 1 day, then 3, then 7, then 21 days. A card that you miss starts again at the first step.
How to use the worksheets
Each lesson has one or more worksheets. A worksheet is a small calculator with an explanation that changes with the numbers. Use each worksheet like this:
- Read the title and the first sentence. They tell you what the worksheet shows.
- Before you change a number, say what you think will change.
- Change one input at a time.
- Read the gray explanation under the numbers. It changes with each input.
- Try the extremes: a very small value and a very large one.
You can type short values in a number field, for example 1.5M, 250k, or 20%. Each slider also moves with the arrow keys. Many calculators in the course also have their own page in the calculators list, for when you want the tool without the lesson.
Other parts of the course
The course has three more tools:
- The glossary. A word with a dotted underline, such as dilution, has a definition. Put the pointer on the word to read it, and click the word to open the glossary. On a phone, tap the word one time to read the definition.
- The review deck. The review deck shows the terms of the lessons that you completed, on the schedule above.
- Your progress. The button at the end of each lesson marks the lesson as complete. Your progress and your quiz scores stay in your browser. Nothing goes to a server.
How this course treats facts
Startup funding changes fast, so the course is strict about its numbers:
- Each market number has a period, for example “In Q2 2026, the median seed round was …”. Each lesson lists its sources at the end.
- When two data sources give different values, the lesson shows the two values and tells you why they are different.
- The companies are invented. Northwind Labs, with its founders Ada and Ben, is in examples across the course.
- Legal and tax rules have a date. Speak to a lawyer or a tax professional before you act on one.
- The course does not sell a product, and it names the conventions behind each calculation.
The next lesson, the whole story in one sitting, follows Northwind from two founders to a sale. It gives you a map of the full course.
Check yourself
These four questions are about this lesson.
Quiz · question 1 of 4
Test what you learned
You have one month to learn SAFE math. Which study plan gives the best memory at the end of the month?
Pretest: what do you know today?
This pretest has 10 questions from across the course. Most of them are about lessons that you did not read yet, so expect some wrong answers. Wrong answers are also useful. A test before a lesson makes later learning better, if feedback comes after it (Richland, Kornell, and Kao, 2009).
Do not look up the answers. Guess, then read the feedback for each option. The capstone module ends with a final exam of the same type, so that you can compare your two scores.
Quiz · question 1 of 10
Pretest: 10 questions from the full course
You hold 5,000,000 of the 10,000,000 shares of a company. The company issues 2,500,000 new shares to an investor. What part of the company do you own now?
Your score today is a start point, not a grade. Each question that you missed has a lesson in this course that answers it.
Key takeaways
- 1.Each lesson uses one loop. You answer a question, make a prediction, read short explanations, move the numbers, do problems, and take a quiz.
- 2.A self-test helps memory more than a second read. In one study, students who tested themselves remembered 61% after one week, against 40% for students who only read.
- 3.A wrong guess is useful when feedback comes after it. Make the guess, then read why the correct answer is correct.
- 4.Spaced reviews keep ideas in memory. The review deck shows a term that you know again after 1, 3, 7, and 21 days.
- 5.The numbers in this course have a date and a source, and the companies are invented. Speak to a lawyer or a tax professional before you act on a legal or tax rule.
Sources
- 1.Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention. Roediger and Karpicke, Psychological Science 17(3), 2006.
- 2.Improving Students' Learning With Effective Learning Techniques. Dunlosky et al., Psychological Science in the Public Interest 14(1), 2013.
- 3.Active learning increases student performance in science, engineering, and mathematics. Freeman et al., PNAS 111, 2014.
- 4.Spacing Effects in Learning: A Temporal Ridgeline of Optimal Retention. Cepeda et al., Psychological Science 19, 2008.
- 5.Testing effect (pretests, and Butler and Roediger, 2008). Wikipedia, Accessed 2026-11-04.
- 6.Worked-example effect. Wikipedia, Accessed 2026-11-04.
- 7.Structuring the Transition From Example Study to Problem Solving. Renkl and Atkinson, Educational Psychologist, 2003.
- 8.The Power of Feedback. Hattie and Timperley, Review of Educational Research 77(1), 2007.
- 9.The effects of feedback interventions on performance. Kluger and DeNisi, Psychological Bulletin 119(2), 1996.
- 10.Communicating with Interactive Articles. Hohman et al., Distill, 2020.
- 11.Explorable Explanations: 4 More Design Patterns. Nicky Case, 2018-06-20.
- 12.A Trainable Spaced Repetition Model for Language Learning. Settles and Meeder, ACL, 2016.