scientific-decision

Scientific Decision-Making and the Third Alternative

R — Original Text (Reading)

Step 1: Stay focused on the core goal, filter out most problems, and do not get entangled in trivial matters; Step 2: When facing key problems, do not get stuck on the only available options; find the core goal, transform the problem into a short-answer question, and seek a third alternative; Step 3: Bypass cognitive biases, activate System 2, and enter slow thinking; Step 4: Establish criteria, score all the options, and quantify the results to compare their magnitude. — Chapter 21, "How to Make Scientific Decisions"

I — Methodological Skeleton (Interpretation)

1. The four steps of scientific decision-making are a funnel from coarse to fine: first cut out unimportant matters, then break through the option framework, then eliminate biases, and finally quantify
2. Step 1: Stay focused on the goal, filter out trivial matters. Most dilemmas are not worth agonizing over; focus on the core goal
3. Step 2: Find a third alternative. When facing a multiple-choice question, do not look only at the options themselves—restore the goal → turn the multiple-choice question into a short-answer question → find a win-win solution
4. Step 3: Bypass cognitive biases. Three major biases: reflection effect (betting bigger when facing losses), certainty effect (seeking security when facing gains), anchoring effect (being anchored by a preconceived numerical value)
5. Activate "System 2" (slow thinking), and use rational analysis to replace intuitive reactions
6. Step 4: Quantified decision-making. Assign weights → score item by item → weighted sum → let the numbers speak
7. 37% rule (wheat ear theory): For a one-time choice with no going back, spend the first 37% of the time only observing and not choosing to establish a benchmark, then after that immediately act on any option that exceeds the benchmark
8. The essence of the third alternative is to step outside the "either-or" framework and find a new solution that maximizes the interests of both sides

A1 — Application in the Book (Past Application)

• Case of choosing a husband: Use a quantified decision table, assign weights to the three major categories of "past, present, future" (such as 30%/50%/20%), score sub-items under each category, and finally use numbers to replace agonizing
• Mate selection by wheat ear theory: When choosing a partner from age 22 to 30, spend the first 3 years (37%) only observing and not choosing to establish the highest-score benchmark, then in the following 5 years immediately act on anyone who exceeds the benchmark
• Third alternative—two swords: Choose the sharp one or the good-looking one? Restore the goal (going to the battlefield or serving as decoration); once the goal is determined, the answer appears automatically
• Cognitive bias—shopping: Two pieces of clothing at the same price of 399 yuan, one originally priced at 399, the other originally priced at 3999 at a 10% discount; people think the second one is cheaper—this is the anchoring effect

A2 — Trigger Scenarios (Future Trigger)

When to use:

• Facing a major life/business choice and feeling extremely torn
• Two options both seem unsatisfactory, but it feels like "one must be chosen"
• Worried that your judgment is being misled by emotions or intuition
• Need to prove to others the reasonableness of a decision (rather than making a snap judgment)

Language signals:

• "Torn between choosing A or B"
• "Neither is good, but I have to choose one"
• "Help me make a decision"
• "How can I choose so I won't regret it?"
• "I think this choice is important"

Distinguish from adjacent skills:

• vs. structured-thinking: structuring is "how to break down the problem"; scientific decision-making is "how to make the optimal choice from the options"; structuring is the prerequisite step of decision-making
• vs. systems-thinking: systems thinking analyzes "why it is this way"; scientific decision-making determines "what to do next"

E — Executable Steps (Execution)

Step 1: Restore the goal, judge the level of the problem

• Write down the core goal you truly want to achieve (not the options themselves)
• Judge whether this problem is worth spending time on for decision-making—80% of small matters are filtered out directly
• If it is a key problem, proceed to the next step
• Completion standard: Be able to write the core goal in one sentence, and confirm that "this problem is worth deep decision-making"

Step 2: Break through the options framework, look for a third choice

• List the options you currently face, and ask yourself: "Who gave me these options?"
• Return to the core goal, and transform the multiple-choice question into a short-answer question: "To achieve the goal, what else can I do?"
• Look for a third solution that can maximize the interests of the relevant parties
• If it involves a choice that cannot be reversed, apply the 37% rule: for the first 37% of the time, only observe and build a baseline
• Completion standard: Produce at least one new solution beyond the original options, or confirm that a baseline has been established using the 37% rule

Step 3: Bypass biases, quantify and score

• Self-check three major biases: Am I taking risks because I fear loss (reflection effect)? Am I giving up a better option because I crave certainty (certainty effect)? Am I influenced by some anchor value (anchoring effect)?
• Build a quantitative table: list evaluation dimensions → assign weights → score each option item by item (0-100)
• Weighted sum, compare magnitudes using numbers
• Completion standard: Each option has a weighted total score, and the highest score is the rational optimal choice

B — Boundary

Scenarios not to use:

• Daily low-risk small decisions (what to eat, what to wear); quantitative decision-making wastes time and energy
• Scenarios requiring extremely fast reactions (crisis handling); there is no time to go through the four-step process
• Purely subjective aesthetic personal preference issues

Failure modes:

• Over-quantification: forcibly scoring dimensions that should rely on intuition; the numbers look precise but are actually false
• Weight-setting bias: the assignment of weights itself reflects preferences; if preferences are affected by bias, quantification is meaningless
• Analysis paralysis: after completing the four steps, still not daring to choose, falling into an infinite loop

Author blind spots:

• Lack of a self-check mechanism for "how to determine whether the weight assignment is reasonable"; the weights themselves are the biggest subjective judgment point
• The 37% rule assumes that evaluation targets appear randomly, but in reality people can change the probability distribution through active search

Related skills

This skill is related to the following skills:

• metacognition-control(Prerequisite dependency): Scientific decision-making requires activating "System 2" for slow thinking, and metacognitive control is exactly the switch that helps you shift from "brain autopilot" to "rational takeover," and is a prerequisite for rational decision-making.
• mirror-world(Prerequisite): The four-step process of scientific decision-making is most effective in the "control domain" and "probability domain" of the mirror world. First use the mirror world to determine which domain an event falls into, then decide whether the scientific decision-making process applies.