Three minutes on the machinery. Why the course is built in slides, why you are asked to guess before you are told, and why you are never blocked from moving on.
F0.1.1 Why you are asked to guess first
Every screen ends in questions, and some of them arrive before the answer does.
The course is built in screens. Each screen holds one idea, then a few questions about it. Some of those questions land before you have been told the answer, and that is the part worth explaining first.
Guessing wrong, then being told, beats reading the answer cold
A 2025 review pulled together sixty years of research on this. Material that people had been questioned about before reading it came back to them later. Material they had only read did not.
The measured effect was 0.66 for the questioned material and 0.01 for the rest. On that scale, anything around 0.4 is usually treated as worth having, and 0.01 is another way of writing nothing.
So a wrong guess is not a side effect of the method. It is the method. That is why the estimate questions ask you for a number you have no way of knowing yet.
There is one condition and it is strict. The answer has to arrive straight after the guess. In the study that set the boundary, delaying the feedback removed the benefit entirely. Every question here answers itself the moment you commit.
The explanation under the answer is half the point
Feedback that only says right or wrong measures about 0.05, so close enough to nothing. Feedback that explains why measures about 0.49. Same question, and the difference between the two is one paragraph.
Read it when you were wrong. Read it when you were right too, because a right answer for the wrong reason looks identical from the outside. The explanation usually names the tempting wrong answer and says why it was tempting.
| What you see on a screen | What it is doing |
|---|
| One line under the title | Frames the screen in a sentence. Safe to skip. |
| The body | One idea, in prose, tables and lists. Never two ideas. |
| The questions at the end | Practice at pulling it back out, not a test. Nothing is recorded against you. |
| The explanation after each answer | Carries roughly half the value of the question. |
Four parts, on every screen in the course.
Four out of five is the targetQuestions are written so that about four in five are gettable from the screen you just read. That rate is not an accident: the most effective classroom teachers measured were running near 82% correct, and the least effective near 73%. Below about half correct, testing stops doing much at all.
So missing one is the design working. Missing four means the screen failed, not you.
How to work a screen
- Read it once.
- Answer without scrolling back up. Being unsure is the point of the exercise.
- Read the explanation, including when you got it right.
- Move on, even if it does not feel finished.
Covered in Module F1, What beer isThe first real module. Four ingredients, three numbers off a menu, and the one division that explains most of what you will taste. Everything after F1 assumes it.
Check yourself
- Before you read on. A screen ends in five questions. How many is it written for you to get right?
Answer: 4questions. Too easy and there is nothing left to learn. Too hard and people stop. The best classroom teachers land somewhere in between.
Four of five, so about 80%. The most effective teachers measured ran at 82% correct and the least effective at 73%, and below about half correct the whole method stops working. Guessing five is the tempting answer, because a quiz that you ace feels like a quiz that went well. A screen where you miss one is doing its job.
- Why does the course ask you to guess a number before it tells you that number?
Answer: Guessing first, even wrong, makes the answer stick better than reading it cold
The guess creates the gap that the answer then fills. Across sixty years of studies, pre-questioned material measured 0.66 and material that was only read measured 0.01. Time on task is the tempting wrong answer, because effort does feel like the active ingredient. It is not: the same reading, unquestioned, produced almost nothing. Nothing here is checked or recorded either.
- Getting a question wrong before you have been taught the answer is wasted effort.
Answer: False
False. A wrong answer is what makes the right one land, so guessing badly is doing the work rather than failing at it. This one feels wrong because school treated a wrong answer as a cost. The single condition is that the answer arrives immediately, which is why nothing here is held back until the end of a module.
- Which part of a question is worth reading even when you got the answer right?
Answer: the explanation
The explanation. Feedback that only says right or wrong measures about 0.05, and feedback that explains measures about 0.49. Skipping it when you were right is the most common way to waste a question, because a lucky guess and a solid answer look exactly the same on the screen.
- Which of these actually appear on a screen in this course?
Answer: A one line framing sentence under the title · A short body covering a single idea · Questions with an explanation after each one
The first three. There is no score, and the next screen explains why that is deliberate rather than unfinished. The framing sentence is the one you can safely skip. The body is held to one idea on purpose, because breaking material into designed chunks is one of the better supported things anyone can do for a beginner.
F0.1.2 Nothing is locked, nothing is scored
Three rules about moving through the course, and one habit worth more than the three of them together.
You can skip anything
Any question, any screen, any class. There is no gate anywhere in the course and no score to clear before the next thing opens.
Courses that make you pass before continuing do raise the scores of the people who finish. In 23 of 32 studies that tested it, they also raised the number of people who did not finish, and the time it took went up by about 4%. For a course nobody is making you take, that is a bad trade.
Skipped questions are not lost either. They come back later in the review queue, which is the other reason a gate is not needed.
No points, no badges, no streaks
Rewards handed out for finishing a task measurably reduce the reason people do the task at all. The effect across a large body of work is minus 0.36. Feedback that tells you what you just did well goes the other way, plus 0.43 in adults.
So the course spends its budget on explanations instead of trophies. When you get something right it will tell you what you got right, and it will phrase it differently the next time, which is the version of praise that has evidence behind it.
The bar along the top is a map. It says where you are and how much is left, and it does nothing else. In a review of ten common game elements, the progress bar was the one element that showed no learning benefit at all. Survey work separately found that bars which slow down near the end raise abandonment by more than half.
The habit that does most of the work
Come back after a few days and answer the questions again. Do not re-read the screen. That single habit is worth more than everything else on this page.
Leaving a gap before you return measures 0.85, which is one of the largest effects in the whole of the learning research. The useful gap is roughly 10% to 20% of how long you want to keep the material. If you want this in a month, come back after about four days.
Re-reading feels like progress because the words are familiar. Familiar is not the same as available. Answering from memory is the thing that was measured, and it is also the thing you will be doing in a bar, with a menu in front of you and somebody waiting.
Worth rememberingFour days, then answer. Do not re-read.
Four days is the gap. Answering is the method. Re-reading is the comfortable thing that feels like studying and is not. If you only take one instruction out of this module, take this one.
| What a course usually does | What this one does | Why |
|---|
| Locks the next part until you pass | Nothing is locked | Gates raised the number of people who quit in most studies that tested them |
| Points, badges and streaks | None of them | Rewards for finishing measurably reduce the reason to do the thing |
| A progress bar that rewards you | A bar that only says where you are | The one game element with no measured learning benefit |
| Re-read the module to revise | Re-answer the questions after a few days | Pulling it out of memory is what was measured, not putting it back in |
Where you are
Twenty one modules in three parts. The foundations you are in now, then eleven modules of styles, then four on putting it to use. Part 2 assumes Part 1 throughout and says so rather than repeating it.
The forward order is fixed, and that is also a decision. Content broken into a designed sequence measures around 0.32. Let the reader reorder it and that drops to 0.15, which is close enough to nothing. Going back is free, because going back costs nothing.
How much to trust these numbersEvery figure on this page is an average across many studies of many people, and several of them come from classrooms rather than from bars. None of it is a promise about you. The honest claim is narrower: these are the choices with the best evidence behind them, so they are the ones the course made.
Covered in Module F5, The decision treeIf you have five minutes and a menu in front of you. Three questions that end on a style name. Colour, then weight, then what leads.
Check yourself
- You are in Part 1, the foundations. What do the eleven style modules that follow assume?
Answer: That you have done the foundations, which is why they come first
The style modules refer back to the foundations by link rather than re-teaching them, so reading them first costs you the explanations. Any order is the tempting answer, because letting people choose sounds generous. It measures badly: a fixed sequence came out around 0.32 and a reorderable one around 0.15, and learner control on its own is worth close to nothing.
- Sort these into what the evidence supports and what only feels productive.
Answer: Guessing before you are told the answer goes in Backed by the evidence · Answering the questions again a few days later goes in Backed by the evidence · Reading the explanation when you were right goes in Backed by the evidence · Reading the screen a second time straight away goes in Feels productive, and is not · Opening the word bank before you have tried goes in Feels productive, and is not · Keeping a streak of correct answers going goes in Feels productive, and is not
The first three are the whole method: guess, retrieve, and read why. Immediate re-reading is the classic one that feels like work, because familiarity is easy to mistake for knowing. The word bank is the sneaky one. Being given the options turns a recall task into a recognition task, and in one study a constrained guess scored 39% against 57% for no prompt at all. Ask for the bank when you are stuck, not before.
- You come back after four days and get several questions wrong that you got right first time. That means the first pass failed.
Answer: False
False, and this is the moment most people quit a good habit. Some forgetting between sessions is the condition the spacing effect needs. A gap you sail through was too short to do anything. The second answer, dug out of a half faded memory, is the one that lasts.
- Before you read on. You want this material to still be there in a month. How many days after a class should you come back and answer again?
Answer: 4days. The rule of thumb from the spacing research is 10% to 20% of how long you want to keep it. A month is the target here.
About four days. Ten to twenty percent of a month lands between three and six. Most people guess a week or two, because a longer gap feels more serious. Too long is a real cost: past a certain point you have forgotten it completely and you are learning it from scratch rather than reaching for it.
- Pair each situation with the move the course is built for.
Answer: A question you cannot answer with Guess, then read the explanation · A screen that is not landing with Skip it. Nothing is locked · A class you finished on Monday with Answer it again on Friday, without re-reading · A question you got right with Read the explanation anyway
All four are the same instruction seen from different angles: keep moving, and get the material out of your head rather than back into it. The one people resist is skipping, because skipping feels like quitting. Gates were tested and they pushed more people out of courses than they helped through them.