In 2006, a psychologist at Washington University did something almost cruel to his students. Henry Roediger divided them into groups, gave them all a passage about sea otters to learn, then asked half of them to study it over and over—exactly what you'd do the night before a test. The other half? They had to keep taking quizzes on it without looking at their notes. Five minutes later, Roediger tested everyone. The studiers crushed it. The poor quiz-takers struggled to remember half as much.
Here's the twist: one week later, Roediger tested them again. This time, the quiz-takers remembered 50% more than the studiers. The students who felt like they were failing had actually been building memories that lasted. The confident re-readers? Their knowledge had evaporated like morning fog.
This is the beautiful lie your brain tells you when you study. That comfortable feeling of recognition—I've seen this before, I definitely know it—isn't learning. It's fluency, and it's an illusion. Four remarkable books published between 2014 and 2019 have tried to expose this illusion and teach us what actually works. They don't all agree. Some make claims that crumble under scientific scrutiny. But together, they reveal something counterintuitive: the harder learning feels, the better it often is.
Each of these four books comes from a radically different world, and understanding where they're coming from matters enormously.
Make It Stick (2014) has the most impressive scientific pedigree. Henry Roediger III has spent 50 years studying memory at Washington University's psychology lab. His Google Scholar h-index is 125—a number that makes researchers jealous. He was elected to the National Academy of Sciences in 2017. His co-author Mark McDaniel runs the Center for Integrative Research on Cognition, Learning, and Education. These are not self-help gurus; they're the scientists other scientists cite. The book emerged from a $6.47 million, 11-university research collaboration funded by the McDonnell Foundation. It's sold 1.5 million copies.
A Mind for Numbers (2014) comes from Barbara Oakley, and her story is pure Hollywood. She flunked math through elementary, middle, and high school. She enlisted in the Army, learned Russian, worked as a translator on Soviet trawlers in the Bering Sea, and spent a season as a radio operator at the South Pole. At 26, she went back to school to "retool her brain" for engineering. Now she's a Distinguished Professor and creator of "Learning How to Learn," a Coursera course with over 4 million students—making it one of the most popular online courses ever created. She co-developed it with Terrence Sejnowski, one of only ten people who holds awards from all three U.S. National Academies.
Ultralearning (2019) is different. Scott Young isn't a scientist—he's a blogger who became famous through spectacular self-experiments. In 2012, he claimed to complete MIT's 4-year computer science curriculum in 12 months using free online materials. In 2014, he spent a year traveling to Spain, Brazil, China, and Korea while refusing to speak English. These stunts generated TEDx talks, media coverage, and a devoted online following. But Young himself admits he would "completely rewrite" his earlier work and has publicly revised some of his stronger claims.
How to Take Smart Notes (2017) comes from Sönke Ahrens, a German lecturer in philosophy of education who became fascinated with Niklas Luhmann. Luhmann was a sociologist who published over 70 books and 400 scholarly articles using a method called Zettelkasten—a box of 90,000 interconnected index cards that he called his "communication partner." The book has sold 100,000 copies and sparked a movement of "personal knowledge management" tools like Roam Research and Obsidian.
Despite their different origins, these books converge on several core truths that have held up remarkably well under scientific scrutiny.
Testing yourself beats re-reading. This is the testing effect, and it's about as solid as psychological research gets. Meta-analyses pooling hundreds of studies find effect sizes around d = 0.5-0.7—meaning students who self-test typically score about half a standard deviation higher than those who just review. Make It Stick is built on this foundation. A Mind for Numbers calls it "recall practice" and builds flashcard strategies around it. Ultralearning devotes an entire chapter to "retrieval" as one of its nine principles. Even How to Take Smart Notes incorporates it by forcing you to generate notes in your own words.
Spacing beats cramming. The spacing effect is one of psychology's most replicated findings, stretching back to Hermann Ebbinghaus in 1885. A meta-analysis of 839 assessments confirms that distributed practice consistently outperforms massed practice. All four books preach this gospel. Make It Stick explains why some forgetting between sessions actually helps—retrieval becomes more effortful and thus more strengthening. A Mind for Numbers uses the analogy of mortar drying between bricks. Ultralearning recommends spaced repetition systems like Anki for retention. How to Take Smart Notes builds spacing into the process of revisiting and connecting old notes.
Mixing topics beats drilling one thing. Interleaving—alternating between different types of problems rather than practicing one type repeatedly—gets strong support from all four books. Meta-analyses find an overall effect of g = 0.42, though it works better for some domains (perceptual learning, math) than others (vocabulary). Make It Stick features this prominently. A Mind for Numbers warns that drilling one problem type creates the "Einstellung effect"—your brain gets stuck in one approach. Ultralearning calls it experimentation.
You're a terrible judge of whether you're learning. Perhaps the most important shared message is epistemic humility. Re-reading feels productive but isn't. Highlighting feels active but is nearly useless. Fluency breeds false confidence. The subjective experience of learning tells you almost nothing about actual learning. This uncomfortable truth appears in every book.
Barbara Oakley's most famous contribution is the idea of "focused" versus "diffuse" thinking modes. She describes focused mode like a pinball machine with bumpers packed tightly together—thoughts bounce intensely in a small area. Diffuse mode is like a pinball machine with bumpers spread far apart—thoughts wander freely, making unexpected connections.
The underlying neuroscience is real. The Default Mode Network (DMN) is a well-documented brain network that activates during rest and mind-wandering. Task-positive networks activate during focused attention. A 2025 study with 2,433 participants found that creativity correlates with dynamic switching between the DMN and Executive Control Network. The incubation effect—where taking a break from a problem helps you solve it—has been confirmed in multiple meta-analyses.
But here's where popular accounts get sloppy. The brain doesn't actually toggle between two discrete modes like a light switch. The networks interact in complex, dynamic ways. Oakley acknowledges her "pinball machine" metaphor is an "oversimplification," but readers often miss this caveat. The claim that "you can't be in both modes simultaneously" is neurologically imprecise—the reality involves continuous interplay between networks.
Evidence rating: ⭐⭐⭐ Moderate. The underlying neuroscience is real, but the two-mode framing is pedagogically useful rather than literally accurate.
How to Take Smart Notes rests on a singular example: Niklas Luhmann, a German sociologist who maintained a box of 90,000 interlinked index cards and produced one of the 20th century's most prolific academic outputs. The method seems magical—an external brain that thinks alongside you.
But here's the uncomfortable truth: no one has ever run a controlled study testing whether Zettelkasten actually works better than other note-taking methods. The entire evidence base is one extremely productive German professor. That's not nothing—but it's also survivorship bias. We never hear about the hundreds of academics who tried elaborate note systems and produced nothing remarkable.
The component strategies have support. Elaborative interrogation (asking "why" and "how" as you read) improves learning. Writing in your own words beats copying verbatim. Making connections between ideas aids memory. But the specific claim that Zettelkasten's architecture is superior to, say, a well-organized notebook or a standard outline? Completely untested.
Evidence rating: ⭐⭐ Weak. The component techniques are supported; the system as a whole is anecdotal.
Scott Young's Ultralearning contains the most dramatic claims and requires the most skeptical reading. His MIT Challenge—learning a four-year computer science degree in 12 months—is inspirational storytelling. But what did it actually prove?
Young didn't attend MIT. He didn't interact with professors or classmates. He watched lectures at accelerated speed and passed final exams using MIT's freely available materials. He demonstrated that a disciplined, intelligent adult could teach himself undergraduate computer science from existing resources. That's impressive! But it's not the same as compressing four years of education into one, and Young has acknowledged the "simplifications" in his original framing.
More importantly, the book relies heavily on case studies of exceptional performers: Eric Barone (who spent five years teaching himself game development to create Stardew Valley), Benny Lewis (the polyglot behind "Fluent in 3 Months"), and Roger Craig (who used data analysis to dominate Jeopardy!). These are fascinating stories. They're also textbook survivorship bias—we're examining the winners without seeing the many people who tried similar intense approaches and failed.
Young's core scientific citations—retrieval practice, spacing, deliberate practice—are well-supported. But his stronger claims about intensity, time compression, and career acceleration lack rigorous evidence. Young himself now says he "over-emphasized practice and de-emphasized examples and explanations" and trusts actual researchers "more than me."
Evidence rating: Mixed. The cited research is solid; the specific intensity claims are anecdotal.
Speaking of oversimplification, one concept that appears across these books deserves a reality check: deliberate practice.
Anders Ericsson's research on deliberate practice—focused, effortful training with feedback—is foundational to Ultralearning and appears in A Mind for Numbers. But the popularized version, especially the "10,000 hours to expertise" claim (spread by Malcolm Gladwell, not Ericsson), has not held up.
Meta-analyses find that deliberate practice explains only 18-26% of variance in performance, not the near-total explanation implied by pop-science accounts. In chess, some players reached master level in 3,016 hours while others never got there despite 25,000+ hours of practice. The 10,000 figure was an average, not a threshold—and averages hide enormous individual variation based on starting age, innate abilities, and other factors.
Evidence rating: ⭐⭐⭐ Moderate. Practice matters and matters a lot. But the strong claims in popular books exaggerate its explanatory power.
These four books target different readers and serve different purposes.
Make It Stick is the most scientifically rigorous and the best choice for educators, parents, or students who want to understand the research directly. Its authors literally did the experiments. The writing can be "tedious and repetitive" (a common criticism), but the information density is high and the claims are trustworthy.
A Mind for Numbers is ideal for students who've struggled with technical subjects and need both encouragement and technique. Oakley's personal transformation story creates emotional resonance, and her techniques (Pomodoro, diffuse-mode activation, chunking) are immediately actionable. Just remember the neuroscience framing is somewhat simplified.
Ultralearning suits ambitious self-directed learners who want to tackle aggressive projects—learning a language, picking up programming, mastering a creative skill. The frameworks are useful for planning. But read it as inspiration and structure, not as scientific proof that his specific approach works better than alternatives.
How to Take Smart Notes is for knowledge workers, academics, and writers who deal with large volumes of information over long time horizons. If you're writing a thesis, building a research program, or synthesizing ideas across years, Zettelkasten might genuinely help. But if you're a student cramming for finals or a casual learner, this system is likely overkill.
If we rank the core claims from these books by scientific support, a clear hierarchy emerges:
| Technique | Evidence Quality | Key Finding |
|---|---|---|
| Retrieval practice (testing) | ⭐⭐⭐⭐⭐ Excellent | Effect sizes d = 0.5-0.7; replicated across labs and classrooms |
| Spaced repetition | ⭐⭐⭐⭐⭐ Excellent | 140 years of consistent research since Ebbinghaus |
| Interleaving | ⭐⭐⭐⭐ Strong | Works especially well for perceptual categories and math (g = 0.42) |
| Desirable difficulties | ⭐⭐⭐⭐ Strong | Framework well-supported; component techniques proven |
| Focused/diffuse thinking | ⭐⭐⭐ Moderate | Real neuroscience, oversimplified metaphor |
| Deliberate practice | ⭐⭐⭐ Moderate | Effect real but smaller than popularized |
| Zettelkasten system | ⭐⭐ Weak | Components supported; specific system untested |
| Ultralearning intensity claims | ⭐⭐ Weak | Anecdotal; time compression claims unverified |
Here's the counterintuitive insight that ties everything together: the interventions that work best—testing yourself, spacing out practice, mixing up topics—all feel worse in the moment than the comfortable alternatives.
When you quiz yourself and struggle to remember, it feels like failure. When you wait a week between study sessions and half-forget the material, it feels inefficient. When you mix problem types and keep getting confused, it feels unproductive. Meanwhile, re-reading your notes feels smooth. Highlighting feels active. Cramming the night before feels intense and memorable.
Your feelings are lying to you. The smooth path builds fragile knowledge. The rough path builds durable knowledge.
This is what Robert Bjork called "desirable difficulties"—challenges that slow initial learning but enhance long-term retention. Not all difficulties are desirable (struggling with incomprehensible material doesn't help), but the right kind of productive struggle is precisely what builds lasting memory.
So what should you actually do? Here's a synthesis that stays true to the evidence:
Test yourself constantly. Don't just reread—close the book and try to recall what you learned. Use flashcards, practice problems, or blank paper where you write everything you remember. This is the single most evidence-backed intervention.
Space your study. Review material at expanding intervals: today, tomorrow, next week, next month. Let yourself forget a little between sessions—that forgetting is what makes retrieval effortful and effective.
Mix it up. Don't drill one topic or problem type until you've "mastered" it. Interleave different subjects or problem types in a single session. It will feel harder. That's the point.
Embrace productive struggle. If practice feels too easy, you're probably not learning much. Seek the sweet spot where you're challenged but not overwhelmed.
Take breaks for insight. When stuck on a hard problem, step away. Walk, shower, sleep on it. Your diffuse networks might find connections your focused attention missed—though don't expect magic.
Connect ideas to each other. Whether you use Zettelkasten or just a notebook, actively link new ideas to things you already know. Elaboration strengthens memory.
Be humble about what you know. Use objective feedback (tests, quizzes, practice problems) rather than subjective feelings to assess your learning. You cannot trust your intuitions about what you've mastered.
These books have sold millions of copies and reached millions of students. They've done genuine good by translating cognitive science into practical advice. But they're also products—marketed, packaged, and sometimes oversimplified for broad appeal.
The core findings about retrieval, spacing, and interleaving are rock-solid. The neuroscience metaphors and personal productivity systems are more uncertain. The dramatic claims about extreme learning feats are often anecdotal. The specific systems (Zettelkasten, ultralearning projects) have not been rigorously tested against alternatives.
Read these books, but read them critically. Use the techniques that have strong evidence—testing, spacing, interleaving, active retrieval. Be more cautious about the elaborate systems and intensity claims. And remember: the goal isn't to optimize your study techniques until they're perfect. The goal is to learn things that matter and remember them long enough to use them.
The beautiful lie your brain tells you—that smooth, comfortable learning is working—is still a lie. But now you know how to recognize it.