The Mind Playbook

The whole volume, in one week

A week with the pages

What the evidence supports at each point of an ordinary week. Each box points to the card that backs it and the free tool that does it. Green boxes rest on the volume's best-supported cards.

From the first evening's mark to the re-read after a wait

None of this is therapy. Two of these practices carry warnings the record put there, and the warnings are in the boxes. Pick one or two; nobody in the book did all of them.

Every eveningOne mark on one line
- One named behavior for the whole week
- Mark the day in pencil, sum the week in pen
- Count conduct you can see, like what you ate or spent
Card 2 · Columns
Four days in a rowFifteen minutes on the hardest thing
- Three or four consecutive days, private, then stop
- The page stays paper: no reader
- Expect a bad hour after
- Not if you are already hurt by it; that needs a therapist
Card 4 · Write
BedtimeAfraid you won't sleep? Try to stay awake
- Lie in bed, in the dark, and try not to sleep
- In the morning, one line: did the bed frighten you less?
- It works on the fear more than the clock
Card 9 · Worse
MorningTwo lines before breakfast
- One thing done as if the doubt weren't so
- One task handed back to whoever it belongs to
- Date them; don't score them
Card 5 · Tasks
When you're stuckList what's holding it before you push
- Driving forces on one side, restraining on the other
- Remove a restraint instead of adding resolve
- Decide out loud, in a group you'll face again
- The scores are your guesses
Card 8 · Field, Wants
After a conversationHow it was said, not what
- One line a day on the pace, the flatness, the lateness
- Catch yourself explaining? Ask what came first, never why
Cards 10, 11 · Manner
After a wait you chooseRe-read as a stranger
- Write without a reader, then wait as long as you choose; Horney gave no number
- Mark the place you flinch, and nothing more
- If the reading turns into a case against you, stop
Card 7 · Alone
Before a courseRead the screen, count the hours
- The trials were run on about 75 hours over eight weeks
- Ten minutes on a phone is something else
- Unwanted effects were counted in 2021; read that first
Card 12 · Basement

Part one · evidence 3 or 4

The best-supported pages

Four practices with a real research literature behind them, or, for the evening mark, behind the tracking under it. Every one of them came out smaller than it was sold, and still worth doing.

Card 4 · Four Days

Cost 2Evidence 4Hype 3

The best record in the book, and a small number

Fifteen to twenty minutes a day, three or four days running, about the most upsetting thing you can write about. Across 146 randomized studies it works, at about 0.15 of a standard deviation. The book keeps it for people who are well.

Illness visits per person, before and after (1986, 46 students)

The shrinking number, same scale (d)

1998, about a dozen studies0.47

2006, 146 randomized studies~0.15

The inventor now quotes it as "about .16." Real, clear of zero, and too small for one person to feel.

1986Four groups, four nights, fifteen minutes a night. Only the group that wrote facts and feelings together held flat on illness visits.

2018On the first theory, that writing sets down a held-in load, its author: "I never found a shred of evidence that supported it."

2006What helped most: three or more sessions, fifteen minutes or more, a private setting, a specific instruction. Close to the 1983 design.

201839 trials, 4,009 people: on depression at follow-up the effect was minus 0.03, indistinguishable from zero.

The four daysFifteen to twenty minutes on each of three or four consecutive days, about the most upsetting thing you can bring yourself to write about. Then stop.

Keep it paperNothing in forty years was improved by the writing being read, stored, scored or shown. Write it, and let it be a page.

Expect a bad hourBlood pressure and bad mood rose after the sessions. That distress turned out to be unrelated to the health result.

Don't do this alone about something that already hurts you: eight patients with PTSD got worse. The version that works for trauma is five sessions with a therapist on one named event, a treatment with its own name. And don't expect a number a doctor reads to move.

Use it: the free toolWrite →A timer and a day count for the four days, with a note on who should wait. The writing never enters the page.

Pennebaker & Beall 1986 · Smyth 1998 · Frattaroli 2006 · Reinhold, Bürkner & Holling 2018 · Gidron et al. 1996 · Season 4

Card 2 · The Columns

Cost 2Evidence 3Hype 2

One line a week, one mark a night

Franklin ruled thirteen virtues against seven days, fell far short, and carried the book for life. Nobody has tested the chart. The tracking underneath it holds up.

Franklin's page, 1733: 13 lines, 7 days, 91 cells

Tracking progress toward a goal (2016, 138 studies, 19,951 people)

larger when physically recordedlarger again when reported or public

1733One virtue had the week, the other twelve took their chances. Thirteen weeks a course, four courses a year.

Order"I found myself incorrigible with respect to Order." Scraping wore the paper into holes, so he moved to ivory leaves and pencil.

1784At 78: "fell far short of it, yet I was, by the endeavour, a better and a happier man." The courses lapsed, "but I always carried my little book with me."

1808Lefin copied the chart into Hebrew without naming Franklin, and set down a better rule: daily marks in pencil, weekly sums in pen.

One lineOne named behavior for one week, then the next line. Every modern reprint drops this first by tracking everything at once.

Pencil and penThe day's mark erasable, the week's sum not. The day is where you correct; the week is where you learn.

Count conductWhat you ate, what you spent, whether the papers went back. "Trust the page on Tuesday's conduct. Do not trust it on your soul."

Nobody has run a trial of a virtue chart. The 0.40 is for tracking countable goals. A spot against Sincerity is your own judgment of yourself, and for Humility, as Franklin said, the page can't work at all.

Use it: the free toolColumns →Thirteen rows, seven days, one line in focus a week. It counts weeks kept, not streaks.

Franklin, Autobiography, Part 2 (Passy, 1784) · Lefin 1808 · Harkin et al. 2016 · Season 2

Card 9 · Worse

Cost 1Evidence 3Hype 3

Can't sleep for fear of not sleeping? Try to stay awake.

Frankl's bedtime trick has better evidence than the philosophy it came wrapped in. Its biggest effect is on the fear of the bed, not on the clock.

Ten trials pooled (2022), by what it was compared with

Against nothingLarge improvements on the key symptoms of insomniaGreat reductions in fear of not sleeping (3 studies)

Against other treatmentsSmaller but still moderate, on three outcomesNo better on minutes to fall asleep

The review's effect sizes are not printed here because the book didn't read them. Its authors ask for stronger studies.

The first fair test, 1979: 25 enrolled, 17 analyzed

The instruction beat a placebo and a waiting list. A 2010 review counted who was actually analyzed and found allocation unclear and no blinding.

1939Frankl later dated the technique to 1939, before the camps. It was tested decades later by people with no stake in his philosophy.

CaseA young doctor afraid of sweating was told to show people how much he could sweat: ten quarts at least. The sweat declined to come. One case, told by its author.

GapBetween 2006 and 2018, papers on it for insomnia were "non-existent." There is no product in lying in the dark.

2025A 2025 review of the philosophy itself: 35 studies, d 1.885 against waitlist, nine tenths of the spread between studies not chance. "It is a warning light."

TonightLie in bed, in the dark, and try to stay awake. Make it your job not to sleep.

Tomorrow morningOne dated line: what you did, and whether the bed frightened you less. Not a sleep score.

Who it's forThe person who can't fall asleep because they are afraid of not falling asleep. That fear is its target.

Ten trials over forty years, the best of them small and none well blinded. Against real treatments it did no better on the minutes it takes to fall asleep, the very thing its name promises. The fear result rests on three studies.

Use it: the free toolWorse →Name the thing you dread, get tonight's instruction, and tap one answer in the morning. No scale.

Frankl 1960, 1975 · Ascher & Turner 1979 · Jansson-Fröjmark et al. 2022 · Cornelius-White et al. 2025 · Season 9

Card 12 · The Basement

Cost 3Evidence 3Hype 3

Eight weeks, about seventy-five hours, and a screen to read first

The course has 47 trials behind it and a modest effect. It is also the one practice whose unwanted effects were counted with an instrument built for the purpose, in a close relative, about four decades after it began.

Effect with its interval at eight weeks, and anxiety at 3 to 6 months (2014, 47 trials, 3,515 people)

Lasting effects, longer than (2021, of 78 people)

A day14.1%

A week9.0%

A month6.4%

Three thresholds, not a range. The sample took a close relative of this course, mindfulness-based cognitive therapy.

1979Began in September in Worcester as a ten-week "Stress Reduction and Relaxation Program." The 1982 paper: 51 pain patients, and "uncontrolled" is his own word.

CostAbout 31 hours of class plus a minimum of 42 to 48 hours of home practice: at least 45 minutes a day, six days a week, for eight weeks. "Call it seventy-five hours."

2014"No evidence that meditation programs were better than any active treatment." Only 9 of the 47 trials reported on harms at all.

2021A 44-question interview by an independent assessor, 96 people: 83% reported some side effect, not all negative. 37%, 30 of 81, reported one that affected functioning, at rates similar to other psychological treatments.

Before week oneRead the adverse-effects screen, and price the course: at least 45 minutes a day, six days a week, for eight weeks.

The real doseThe trial numbers belong to the whole course. Ten minutes a day on a phone "is describing something else."

If it goes strangeThat is a documented feature of the practice, not a failure of yours, and there are now people whose work is helping with it.

About a third of a standard deviation, fading by a few months, with depression's interval touching zero. Not shown to beat exercise, muscle relaxation or group cognitive therapy, and most of those comparisons were graded insufficient.

Use it: the free toolBasement →The eight weeks at the published lengths, with the adverse-effects screen before week one.

Kabat-Zinn 1982, 2011 · Goyal et al. 2014 · Britton et al. 2021 · Season 12

Part two · evidence 2

Useful pages, small studies

Each of these has a real finding under it, and most carry a much bigger claim on top. The page costs a notebook. The claim is where to be careful.

Card 8 · The Field

Cost 1Evidence 2Hype 3

Take away what's holding it before you push harder

Kurt Lewin's real finding was one sentence and a wartime experiment with six small groups of women. The scored diagram with his name on it came later, and its numbers are yours.

Served kidney, heart or brains the next week (1943)

Decided as a group52%

Heard a lecture10%

First-timers, group29%

First-timers, lecture0%

23 of 44 women against 4 of 41. Six groups of 13 to 17, checked once, no long follow-up.

Two ways to move a settled level (1947)

Add pushThe line moves, and the old resistance shoves back."Relatively high tension"

Remove what resistsThe line moves with less pushing back."Lower tension"

1943The government wanted families to eat organ meats. Lewin found the "gatekeeper," the person who decides what reaches the table, and tested lecture against group decision.

1947Change comes "either by adding forces in the desired direction or by diminishing opposing forces," and the two leave different tension behind.

131Unfreeze, change, refreeze: in his 1951 book the steps fill 131 words of 338 pages, and he never wrote "refreezing." Festinger, Lippitt and Schein built the model later.

1973Encounter groups, descendants of a door he left open: about a third improved at six months, and 8%, one in twelve, were psychologically injured.

When stuckBefore you add resolve, list what is resisting and ask which of it could simply be removed.

DecidingDecide it out loud, in a group you will face again. That is the one use of "freezing" he actually made.

The two columnsDriving forces on one side, restraining on the other, as a list. Draw it if it helps you think; don't cite it as a finding.

No study shows that scoring the diagram improves a decision. The numbers beside the arrows came out of the head holding the pen. "Remove the two strongest restraints" is a later workshop rule, not Lewin's.

Use it: the free toolField →Driving and restraining forces, scored by you and labeled as yours, then his instruction: take restraints away before you push.

Lewin 1943, 1947 · Cummings, Bridgman & Brown 2016 · Lieberman, Yalom & Miles 1973 · Season 8

Card 7 · Alone

Cost 2Evidence 2Hype 2

Write it, wait, and read it back as a stranger

Karen Horney's method for doing analysis alone came with its ceiling printed in the last chapter. The re-read page shows what you are willing to see, and that moves over months.

The same dated page, two readers

The observerCurious. Finds the place you hurry or explain, marks it, asks what it protects.Keep this reader

The prosecutor"The tyranny of the should." Reads the page as a charge sheet.Stop re-reading

A 2009 study of 111 students over three weeks: reflection early predicted rumination later, and the net effect on depression came out near zero.

Distance, not depth (2008, 141 and 328 people)

Watching as an observerLower depressed mood, still there at day 1 and day 7

Reliving from insideDid not hold their mood

1941Her society voted 24 to 7, with abstentions, to strip her standing as a training analyst. Four colleagues and eleven students walked out with her.

1942"The greatest advantage of writing is that it affords the possibility of going over the notes afterward." She gave no waiting interval: only "afterward" and "later."

LimitHer own chapter 11: "There is no such thing as a complete analysis." In 1943 her institute took a member's training status, and colleagues and her daughter left.

TrialsTwenty trials of written self-examination: about 5% better than control on symptoms, with the anxiety and depression intervals touching zero. Nothing tests her method itself.

WriteFreely, and for no reader, including yourself later. A page that knows it will be re-read starts to perform.

WaitChoose your own interval and admit it's yours, since she gave none. The wait buys distance.

Mark the flinchThe place you skip, hurry or explain. Mark it, ask what it protects, and do nothing else to it.

If the re-reading turns into a case against you, stop; that is the method reaching its limit. "If the charge sheet keeps writing itself without the page, the word to look for is therapist." No score, type or trait comes out of this page.

Use it: the free toolAlone →Write, lock it for the days you choose, then re-read for the flinch. One mark, nothing computed, and a stop rule.

Horney 1942, 1950 · Paris 1994 · Trapnell & Campbell 1999 · Kross & Ayduk 2008 · Takano & Tanno 2009 · Season 7

Card 1 · The Grid

Cost 2Evidence 2Hype 3

Your own words, three people at a time

George Kelly's grid is the only personality instrument the book knows of where every word comes from you. Its arithmetic is honest. The claim that it measured you is not.

One triad: how are two alike and the third different?

The therapy it belongs to, pooled (2007, 22 studies)

Vs no treatment, end0.38

Vs active treatment, end0.21

Vs no treatment, follow-up0.22

Vs active, follow-up0.13

Against active treatment, not significant. A second 2007 review, 27 studies and 1,511 people, found 0.34 against nothing and no significant difference against other therapies.

1955Kelly's two volumes. The one idea that travels: "one does not have to disprove one proposition before entertaining one of its alternatives."

1972Britain turned the grid into a printout: Slater's INGRID computed its main dimensions, and later Shaw's FOCUS clustered the raw ratings.

1980Shaw on the math's equal-spacing assumption: testing it "is an impossible condition to satisfy with a single elicited grid."

2007Change showed least on measures built from the client's own content, which is the grid.

An afternoonTake the people who matter to you three at a time. Ask how two are alike and the third different, and write down your word.

The computed pageRate the cells and see which of your words travel together. Read it as a portrait of your vocabulary on one afternoon.

Another accountTry a new account of yourself without first disproving the old one. That sentence needs no grid.

The words you get shift with which three people you pick, and two people's grids can't be compared. The distance between your self and ideal self on a printout has not been shown to mean anything outside the grid. "Believe the arithmetic and doubt the noun."

Use it: the free toolGrid →Name your people, sort them in threes, rate the cells. It shows which of your words move together and prints no type, score or label.

Kelly 1955 · Shaw 1980 · Holland, Neimeyer, Currier & Berman 2007 · Metcalfe, Winter & Viney 2007 · Season 1

Card 3 · The Word List

Cost 2Evidence 2Hype 4

The slow words are real. What they mean isn't settled.

Jung timed a hundred words with a stopwatch and wrote on page eleven that it measured nothing intricate. A century of "detectors" skipped page eleven.

Jung's median answer times, healthy people, in seconds, against his five-second alarm

Educated men1.3

Men1.6

Concrete nouns1.67

Educated women1.7

Verbs1.90

Women2

"Suspicion of a complex"5+

Sex, schooling and grammar move the time as much as a hidden knot would.

The polygraph as a screen (2003 review, 57 studies)

10→spies among 10,000 employees; set the machine to catch 8

1,606→people flagged

1,598→of them innocent

190438 people, 400 words each, "a one-fifth-second stop-watch." Page eleven: "we do not imagine that we have measured any intricate psychological times by this simple procedure." Tired subjects came back the next day.

1910Kent and Rosanoff ran 1,000 normal subjects, then stopped recording the time: its causes were "numerous and complex."

1998The Supreme Court: "There is simply no consensus that polygraph evidence is reliable."

2013The keyboard heir, the IAT, predicted behavior at .14; plain asking did .12. In 2017 it agreed with itself at .54.

The listA hundred words, one at a time, the first word back written beside each.

The slow onesA word that cost five seconds when its neighbors cost one is a word to sit with. Keep it as a question.

RestedJung and Riklin sent tired subjects home till morning, because a tired answer says nothing about the word. Do the same.

A stall can be a secret, a rare word, a tired evening or your second language, and on Jung's own tables those cost the same. No study scoring his seven signs against an outside check turned up. The famous case is one confession.

Use it: the free toolWords →A hundred words, one at a time, timed. It lists your slow ones and refuses to say what they mean.

Jung & Riklin 1904 · Kent & Rosanoff 1910 · National Research Council 2003 · Oswald et al. 2013 · Gawronski et al. 2017 · Season 3

Card 6 · The Four Columns

Cost 2Evidence 2Hype 4

Four questions, one row a day, for the stuck

What do you want, what are you doing, is it working, what's the plan. The worksheet is good. The theory William Glasser later hung on it is wrong about people who are ill.

The four columns, as Wubbolding wrote them down (WDEP)

WWants

DDoing

EIs it working?

PPlan

The book keeps the third column above all the others: it is the one nobody else can fill in for you.

The same sentence, two people

Functioning and stuck"What am I doing today" instead of "what is wrong with me"A lever

Ill, can't want anything"You are choosing to depress"An accusation

1956At a California state school for girls he asked what each wanted, what she was doing, and whether it worked. The good outcomes rest on his own account.

1965Reality Therapy: a procedure with almost no theory in it. The obituaries put sales at a million and a half.

2003His largest claim: "The basic human problem has nothing to do with the structure or physiology of our brain." The record on major depression, bipolar disorder and schizophrenia says otherwise.

TrialsTwo reviews, 21 and 43 studies, both in the movement's own journals, with brief treatments and often inexperienced therapists. No independent review found.

One rowA want you can state, what you actually did about it, a judgment, and a plan small enough to keep. It fits on an index card.

Column threeOver a month the rows show the gap between what you say you want and what you did, and whether it's closing.

Bad month, not bad year"Use it on yourself in a bad month. Do not use it on someone in a bad year."

The page measures a gap in a person who can act. It can't show that everyone can, and "the brain is never the cause" is contradicted by the record. Fill the want column in your own words; the five needs have never been shown to be more than a list.

Use it: the free toolWants →Want, doing, is it working, plan. One row a day, and no sentence about your brain.

Glasser 1965, 2003 · Wubbolding 2000 · Radtke, Sapp & Farrell 1997 · Kim & Hwang 2006 · Season 6

Part three · evidence 1

Keep the page, drop the label

No controlled trial supports any of these three practices as you would do them yourself; the one test of the solo Bracket did not replicate. Each one still costs very little and does something you can see for yourself. The sentences written across the top of them are what to leave behind.

Card 5 · The Tasks

Cost 1Evidence 1Hype 4

A first memory, a week as if, a task handed back

Three small procedures from Alfred Adler's world, with no trial behind any of them. Keep them. Drop the sentence that says you chose your suffering.

Where the pieces actually come from

1911Adler leaves Freud's society

1912The Neurotic Constitution: a life steered by a private "fiction"

1919From about now, child guidance clinics in Vienna; about thirty by 1930

1931"I would never investigate a personality without asking for the first memory"

1934The clinics are closed

1987"Whose task is this?" coined in Tokyo by Noda Shunsaku, from a Dreikurs question

2013A dialogue puts it under Adler's name; over ten million copies

Adler's own tasks, 1931Three: occupational, social and sexual

"Separation of tasks"1987, and Noda says Adler wrote no such thing

1931"Every memory is a memento": the first one is "the beginning of the autobiography he has made up for himself."

1919Young Karl Popper noticed the theory explains a man who drowns a child and a man who dies saving one equally well. As the book puts it, a thing that explains everything has been tested by nothing.

1931"No experience is a cause of success or failure." Said in clinics built to change children's circumstances, by a doctor who then changed them.

NoneNo trial separates Adlerian therapy from other talking cures. No agreed scoring for first memories, no measured effect for acting as if.

The first memoryWrite down the earliest thing you remember and date it. Don't score it. Read it again in a year and see if you'd still pick it.

One week as ifBehave for a week as if the thing you doubt about yourself were not so, and see what the week does.

One task backAsk whose task this is, and hand one back. Credit it to 1987, not to Adler.

"You chose this," said to someone whose suffering is structural, such as poverty, illness or a harm still going on, is an accusation. Look first at whether the structure can be changed. That is what Adler did.

Use it: the free toolTasks →Two dated lines a day: one thing done as if, one task handed back. Kept, and never counted.

Adler 1912, 1931 · Popper 1963 · Noda Shunsaku, 2016 and 2017 posts · Season 5

Card 10 · The Manner

Cost 1Evidence 1Hype 4

Listen to how it was said

Wilhelm Reich's 1933 instruction shaped a profession before he went somewhere it couldn't follow. The instruction survived. The muscle map has no study behind it, and the energy and the box have no finding.

The instruction from page 213

The WhatThe content: every interpretation understood and acceptedNothing changed

The HowA flat, dying monotone, named out loudWhere the rest was hiding

The nearest measured relative: body psychotherapy (2021, 18 trials, 1,297 patients)

Symptoms, vs mostly waitlists0.56

Distress0.52

At about 7 months0.41

Character analysis itselfnone

None of the 18 trials was character analysis. Against active controls the advantage shrank, and the authors could not rule out bias.

1933Page 57: rather than exhort, wait "until you understand why the patient behaves so and not otherwise," down to why he tells "only every third thought."

1933The movement's own press cancelled it, then sold the book anyway on a 25% commission: seventy copies by October.

1941Einstein's letter explained the "energy" as an ordinary gradient: air cooler at the floor than the ceiling. Reich answered in 25 pages.

1956About six tons of his books and papers were burned in New York, the 1933 book among them. Wrong, and it proves nothing about the box.

One line a dayNot what was said, to you or by you, but how: the pace, the flatness, the lateness, the third thought left out.

Notice first"Most of us exhort. The instruction is to notice first."

Skip the body mapThe claim that emotion is stored in muscle and released by pressing on it has no study behind it: unsupported, and never refuted either.

There is no controlled trial of character analysis; its evidence is the testimony of the field that absorbed it. The honest figure for this practice is no figure. Expect a better ear, nothing more.

Use it: the free toolManner →Not what was said, how. One line a day, and the page counts nothing.

Reich 1933 · Sterba 1982 · Rosendahl, Sattel & Lahmann 2021 · Season 10

Card 11 · The Bracket

Cost 2Evidence 1Hype 3

Ask what came first. Never ask why.

The best instruction in this season is free. The bigger claim, that careful questioning of yourself, alone, gets back what really happened, failed its only test.

Noticing a secretly swapped card (2013, 30 people)

Asked to explain1/3

Trained interviewer, 24 trials80%

2021: the interviewer replaced with written prompts, done alone. "The experiment did not replicate." The authors: "training is required."

One word in the question (1974): estimated speed, mph

"Smashed"40.5

"Hit"34.0

"Contacted"31.8

Same accident. In a second experiment, a week later, 16 of the "smashed" group remembered broken glass that wasn't there, against 7 for "hit" and 6 in the control group.

1938Husserl left some 45,000 pages of shorthand. The season argues he never wrote the bracket down as steps a person could follow.

1988Vermersch banned "why": it turns a person toward justification and away from description. Ask what came first, and then.

2006The method's own test of validity is "the quality of contact with the experience," not whether the description is accurate.

1995Two interviews planted a childhood memory of being lost in a mall in 25% of 24 adults. A 2023 replication: 35% of 123.

Caught explainingWhen you catch yourself explaining why, stop and ask what came first instead.

One momentOne specific time, not the general case. Go back through the senses, where you were and what your hands were doing, in order.

Describe, don't explainLearn to hear the difference in your own voice. "Most of what people say about their inner lives is explanation dressed as report."

Whatever you write down is a description made in contact with a memory, which is a different thing from the memory. The 80% came from one lab and thirty people, not independently repeated.

The toolNo tool, on purpose. The only controlled test of doing this alone did not replicate, and its authors concluded training is required. A free page of prompts would imply what the record can't carry.

Vermersch 1994 · Petitmengin 2006, 2013 · Loftus & Palmer 1974 · Loftus & Pickrell 1995 · Sparby et al. 2021 · Season 11

Index

All twelve on one map

Across: what it costs to do. Up: how well the research backs it. Color: how far the popular version runs ahead. Tap a practice to jump to its card.

Almost everything here is cheap, and most of it is thin

4 Strong

4 Four days of writing

3 Moderate

2 Suggestive

1 Triads 3 Word list 6 Four columns 7 Write, wait, re-read

1 Thin

5 First memory, as if 10 How, not what

11 Ask what came first

1 Very low cost

2 Low

3 Moderate

4 High

Hype gap: 2 moderate 3 large 4 extreme

Eleven of twelve sit at cost 1 or 2, and eight of twelve at evidence 1 or 2. Ratings are one reader's judgment on the book's own scales, argued row by row in the Volume Six matrix. They don't rate how well anything works for you.