Research Notes: Thinking, Fast and Slow → Wild Dogs Leadership

Extracted from the full text (~/projects/my-cfo/mycfo/docs/books/text/thinking-fast-and-slow.txt, line refs to that file). Backing for the Course Design doc and course/leadership-soul-scorecard.md.

1. The core mapping: System 1/2 = Jer's head/soul thesis, scientifically stated

• System 1: fast, automatic, always-on, "the secret author of many of the choices and judgments you make" (~496). System 2: slow, effortful, lazy — "although System 2 believes itself to be where the action is, the automatic System 1 is the hero" (~638).
• Skills become System 1 through prolonged practice (~672, ~1151, ~3634): System 1 "executes skilled responses and generates skilled intuitions, after adequate training." Jer's "leadership muscle memory" is literally this.
• Knowing doesn't change the automatic (Müller-Lyer, ~869): "you cannot decide to see the lines as equal, although you know they are." The only defense is trained recognition of risky situations (~903–915).
• Kahneman's own confession (~14621): "System 1 is not readily educable… my intuitive thinking is just as prone to overconfidence, extreme predictions, and the planning fallacy as it was before I made a study of these issues." → The Nobel laureate, after a career studying bias, is unchanged at the System-1 level. THE proof text that information ≠ formation; only slow practice rewires the automatic.
• Smartness ≠ formedness: Stanovich — "high intelligence does not make people immune to biases" (~1645); engaged vs lazy thinking is a separate trait.

2. Expert intuition = the science of "crock-pot, not microwave"

• Intuition is recognition (Simon, ~8196): "The situation has provided a cue… Intuition is nothing more and nothing less than recognition."
• Trustworthy intuition requires BOTH (~8314): "an environment that is sufficiently regular to be predictable" AND "an opportunity to learn these regularities through prolonged practice." Expertise is "a large collection of miniskills" (~8243) taking ~10,000 hours.
• Wicked environments teach wrong lessons (~8331): the typhoid physician who "developed a sense of clinical infallibility" while infecting patients. Positional power is a wicked environment: subordinates' deference is corrupted feedback, so a senior leader's self-trust grows while their accuracy doesn't.
• Power increases trust in gut (~4684): "Merely reminding people of a time when they had power increases their apparent trust in their own intuition." The more senior, the less reliable the self-assessment.
• "Do not trust anyone — including yourself — to tell you how much you should trust their judgment" (~8309).
• Regression-to-the-mean / flight instructors (~6062, ~14982): "we are statistically punished for being nice and rewarded for being nasty" — experience itself teaches leaders the wrong lesson about harshness vs encouragement.

3. Why self-assessment fails (= why the Scorecard can't use self-ratings)

1. Substitution (~3375): hard questions get silently swapped for easy ones; "you may not even notice that you did not answer the question you were asked." "Am I a humble leader?" becomes "Do I feel good about myself right now?"
2. Self-halo + confirmation retrieval (~2798): "Is Sam friendly?" retrieves only confirming evidence; question wording determines the memory search.
3. Cognitive ease feels like truth (~2123): "familiarity is not easily distinguished from truth" — repeated self-descriptions ("I'm a good listener") feel true by repetition alone.
4. Confidence ≠ accuracy (~7291): "declarations of high confidence mainly tell you that an individual has constructed a coherent story in his mind, not necessarily that the story is true." Illusion of validity: knowing the statistic doesn't dent confidence in one's own case (~7282).
5. Availability asymmetry (~4500): spouses' claimed shares of housework sum >100%; "many members of a collaborative team feel they have done more than their share."
6. Self-exemption (~5931, ~6013): people taught the helping-experiment results "learned nothing at all" — they quietly exempt themselves from general findings.
7. The design answer (~6048): "You are more likely to learn something by finding surprises in your own behavior than by hearing surprising facts about people in general. That is why this book contains questions that are addressed personally to the reader."

4. Leadership-specific findings for course modules

• Halo effect on CEOs (~7050): CEO–firm-success correlation ≈ .30; the stronger CEO leads the stronger firm only ~60% of the time. "If you expected this value to be higher… you are prone to overestimate the predictability of the world" (~7071).
• Built to Last / Good to Great critique (~7108): comparison firms were "more or less lucky"; the winners' edge "shrank to almost nothing" post-study. (Course must teach Collins WITH this caveat — intellectual honesty is part of evidence-armored soul.)
• "A few lucky gambles can crown a reckless leader with a halo of prescience and boldness" (~7029). Award-winning celebrity CEOs subsequently underperform (~8991).
• CFO calibration (~9108): 80%-confidence intervals surprised 67% of the time; "the CFOs did not appear to know that their forecasts were worthless" — and admitting calibrated uncertainty would get them "laughed out of the room" (~9141).
• Loss aversion & the hard calls: losses loom ~2× gains; leaders facing a sure loss (kill the project, exit the bad hire) prefer the gamble (one more quarter) — "risk seeking in the domain of losses is a robust effect" (~15593). Deviating from default is commission and "more likely to evoke regret than doing nothing" (~14461) — why drift feels blameless. Sunk cost: the executive who can't cancel because it "will leave a permanent stain on the record" (~12052); boards replace CEOs partly because the new one "does not carry the same mental accounts."
• Endowment/turf: "Imagine we did not own it; how much would we think it is worth?" (~13124) — the turf-protection antidote question.
• Two selves (~13278): experiencing self vs remembering self; "the tyranny of the remembering self" (~13303); "we all care intensely for the narrative of our own life and very much want it to be a good story, with a decent hero" (~13495). Peak-end + duration neglect → leaders chase peaks; faithful decade-long formation has no peaks, so the remembering self undervalues it (~14305). The amnesic-vacation probe (~13580).
• The boss finding (~13766): in experience-sampling, time with the boss was "the only thing that was worse than being alone." A leader IS their team's experienced day.
• Organizations as System 2 (~14647): "Organizations are better than individuals when it comes to avoiding errors, because they naturally think more slowly and have the power to impose orderly procedures… a factory that manufactures judgments and decisions." Shared bias vocabulary → "a direct link from more precise gossip at the watercooler to better decisions" (~14675). Judge decisions "by how it was made, not only by how it turned out" (~14679).
• Premortem (Klein, ~9234): "Imagine that we are a year into the future. We implemented the plan as it now exists. The outcome was a disaster. Write a brief history of that disaster." Counters the suppression of doubt around a leader who has tipped her hand (~9249).
• Decorrelate errors (~2937): before discussion, all committee members write their position privately — "open discussion gives too much weight to the opinions of those who speak early and assertively."
• Mood/embodiment: acting calm and kind produces feeling calm and kind (~1852); money priming reduces helpfulness (~1895); ego depletion — the parole judges at ~65% approval after meals, ~0% before (~1436): under depletion, the leader's formed character is what shows up.

5. Question formats that produce the "aha, I just did that" moment

Design principles (Kahneman's own): (a) items must evoke a fluent, wrong answer (~2227); (b) self-checkable in seconds; (c) personal surprise > general statistics (~6035); (d) show the taker their own two inconsistent answers; (e) fluent formats EXPOSE the bias, effortful/frequency formats TEACH the correction (~2242, ~11469); (f) concrete behavior recall dodges substitution; global self-ratings invite it.