ROI / Decision Evaluation System (Cross-Author Topic Synthesis)

When to open: When you need to make "should I do X" type decisions (changing jobs / buying school district housing / taking orders / getting a raise / investing in projects) / want to train your team's ROI capability / want to know exactly how to use "AI as a measuring stick". This wiki integrates the perspectives of both yitang Truman and Wang Kai.

One-sentence core:Broad ROI (daily decision-making) ≠ Narrow ROI (financial figures). "When in doubt, use ROI"(yitang internal catchphrase) + "AI is a measuring stick, not a fortune teller"(Wang Kai mindset) + "ROI is a measuring stick, adjust 'not quite good enough' to 'extremely good'"(Truman altitude mindset).

0. The two schools of "ROI"

Perspective	Source	Core model
yitang Truman system school	Yitang / Scientific Decision-Making Complete ROI System (4 main lessons + 5 Candy case collections)	5 lessons + triangle (breadth/depth/height) + 13,000+ assignments
Wang Kai AI decision school	Wang Kai's Decision-Making System for the AI Era / Wang Kai's RSI AI Investment Strategy and the "Good Asset, Good Price, Good Volatility" Framework	AI as measuring stick + 5 mindsets + absolute rational thinking as default + RSI three layers

1. Cross-author consensus (5 points)

1.1 Broad ROI ≠ Narrow ROI (exclusive to yitang but agreed by all schools)

Type	Meaning
Narrow ROI	Return on investment / marketing ROI (financial figures)
Broad ROI(focus of this lesson)	ROI used in daily decision-making / business advancement / team management, with dozens to hundreds of times more use cases than narrow ROI

Typical broad ROI decisions:

Should we provide a free lunch policy for employees?
Should we require all staff to clock in?
Should we do personal IP / livestreaming?
Should we hire a technical co-founder?
Should we buy a school district house for our daughter?

1.2 The 3 major inertia traps in decision-making

Shoot-from-the-hip decision-making —— Making a call after only vague thinking
Careless oversight —— Failing to evaluate comprehensively, leaving blind spots
Held back by face —— Failing to make rational decisions, swayed by personal relationships

1.3 AI as a measuring stick (exclusive to Wang Kai but corroborated by Truman)

Wang Kai's RSI AI Investment Strategy and the "Good Asset, Good Price, Good Volatility" Framework §4:For ambiguous scenarios that humans cannot quantify precisely, let AI score them using a stable framework.

Scenario	AI usage	Value
Market risk level	Score on a 1-10 scale	Determine cash ratio
Enterprise intrinsic value	Multi-model synthesis + interval	Decide whether to buy
Single stock risk	Score across 5 short-, medium- and long-term dimensions	Determine position size
Emotional state	"Am I being overly panicked right now?"	Prevent impulsive decisions
Startup project feasibility	Break down the model and let AI interrogate it	Project approval

Analogy:Insurance company actuary. Give AI a stable framework, and it becomes a ruthless actuarial machine.

Key discipline:Consistency > accuracy. Running the same stock multiple times should yield similar results.

1.4 Absolute rationality / thinking as default

Wang Kai's Decision-Making System for the AI Era §1 5 Mindsets:

#	Mindset
1	View the present from the AGI perspective
2	AI is a measuring stick, not a fortune teller
3	Absolute rationality, thinking as default —— Any important decision must follow the slow-thinking path of "writing out the reasoning process"
4	Rules outweigh results
5	Understanding = being able to do it

The core value of AI's thinking mode: forcing "fast-thinking decisions" into "slow-thinking decisions".

1.5 ROI Triangle (Yitang's proprietary holistic framework)

Yitang / Complete ROI System for Scientific Decision-Making (4 core lessons + 5 Candy case collections) §II:

        高度
        ↑
  宽度 ←+→ 深度
        |
       决策

Dimension	Core question	Operation
Breadth	Identify the key input items and return items to find them out and avoid blind spots	List, deduce, build, check
Depth	Strive for precise evaluation; try quantification, go deep with numbers	X-type → Y-type
Height	Fully consider long-term value, opportunity cost, and time window	Case parking lot / Elevate dimension

2. Explicit disagreements (2 items)

2.1 Disagreement 1: Should decisions "wait for the right moment"?

Perspective	Position
Truman(Mandatory aesthetic training for all members)	"Decisions can be made half a beat slower; take a little time to sort out whether it is worth doing. Many things are actually quite easy to judge"
Wang Kai(Waiting > Thinking)	"Waiting > Thinking. It is not hard to think things through clearly; the hard part is waiting for the right price. Any company can wait for a reasonable price, and most can wait for an undervalued price"
Duan Yongping(cited by Wang Kai)	"Being cheap enough is not the key; understanding well enough is the key"

Real disagreement:The objects of waiting differ:

Truman: Wait for "thinking it through" (short-term vs. decision-making)
Wang Kai: Wait for "the right price" (long-term vs. investment)
Duan Yongping: Wait for "understanding well enough" (cognition vs. decision-making)

2.2 Disagreement 2: Should you proceed when the ROI does not add up?

Yitang / Complete ROI System for Scientific Decision-Making (4 core lessons + 5 Candy case collections) Altitude Chapter + ROI Implementation Training Chapter:

Perspective	Position
Truman's altitude mindset	"ROI is a measuring stick, not for evaluating whether something is worth doing, but for adjusting something that doesn't quite work into something that works very well" —— 5 levers to raise value and lower cost
Wang Kai's discipline	Abandon immediately once consensus is formed but the ROI does not add up —— "Strive to reason things out"

Real disagreement:When the ROI does not add up:

Truman's position:Adjust it to work by pulling levers(increase value / reduce cost / change structure / extend time / change scope)
Wang Kai's position:Abandon if the ROI does not add up(Discipline outweighs results)

Complementary conclusion:

Startup / exploration phase → Truman's lever-pulling adjustment (5-axis adjustment)
Investment / rigorous scenarios → Wang Kai's immediate abandonment (discipline)

3. Yitang's four-character formula: List, Deduce, Build, Check (proprietary and complete)

Yitang / Complete ROI System for Scientific Decision-Making (4 core lessons + 5 Candy case collections) §IV Breadth Chapter:

Character	Meaning
List	Strive to calmly list the costs and benefits of the decision
Deduce	Deduce the whole story using one order or one user
Build	Build a MECE framework or formula for rigorous evaluation
Check	Seek expert consultation, or conduct self-check using the "blind spot arsenal"

3.1 3 categories and 9 typical examples of ROI evaluation

Personal Section:

Should I change jobs?
Should I buy a school district house?
Should I go get an MBA?

Team Section: 4. Should we hold weekly meetings? 5. Should we implement position KPIs? 6. Should we provide free lunch to employees?

Company Section: 7. Should we build an IP? 8. Should we do livestreaming? 9. Should we hire a technical co-founder?

3.2 Classic Case: Liang Jing Builds Her Own Factory (Candy)

Detailed Numerical Breakdown:

Current site rent is 70,000/month, rising to 100,000 the year after next
Purchasing a 1000 sqm factory costs 10 million
Relocation cost: 2 million + 500,000
Labor cost (dormitory 300,000/year + turnover appeasement 100,000 + recruitment +50,000/year)
Business loss cost (1–2 months of business disruption)
Total relocation cost ~11 million
Compared with continuing to rent: with the 10 million investment calculated at an annualized return of 2–4%, roughly it breaks even with rental costs in 9–10 years

Long-term perspective: factory value preservation / industry barriers / cash flow pressure.

3.3 Type X vs Type Y Decision-Makers (In-Depth Section)

Type	Characteristics
Type X decision-maker	Gut calls / vague judgment / simple evaluation
Type Y decision-maker	Quantitative calculation / multivariate analysis / precise numbers

Upgrade path: Type X → Type Y, gradually increasing decision depth.

3.4 Altitude Section: 4 Altitude Dimensions

Long-term value(3–5 years / 10 years)
Company perspective(local vs global)
Opportunity cost(the price of doing A instead of B)
Time window(early / late / missed)

Classic Mental Models:

+1 Thinking: put yourself in the boss's shoes
CEO Weekly Report: reverse-sync key decision information
Decision Hill Climbing + Persuasion Iceberg

4. ROI Implementation Training: 5-Step Training Model (Exclusive)

Yitang / Complete Scientific Decision ROI System (4 main lessons + 5 Candy case sets) §VI Implementation Training Section:

1. 有意识(识别决策,4 动作 + 减少过早拍板 3 建议)
   ↓
2. 写初版(画布升级 3 阶:基础/专业/强行定量 + "换不换"思维)
   ↓
3. 细打磨(ROI 是尺子心法,把不太行调到非常行 / 5 杠杆)
   ↓
4. 有共识(共识 4 层 + 决策爬山+说服冰山曲线 + 集体共建 80/20)
   ↓
5. 做复盘(4 层复盘:决策/参数/迁移/能力)

4.1 3 Suggestions to Reduce Premature Decision-Making

Fictional Blue Team(assume the opposing side's viewpoint)
Fifty-Fifty(force at least 50% uncertainty)
Delay the decision(wait for more information)

4.2 4 Progressive Layers of Consensus

Action consensus(what everyone will do)
Judgment consensus(everyone thinks it is right)
Logic consensus(why it is right)
Information consensus(based on what information)

4.3 ROI Is a Ruler (Key Mental Model)

"ROI is not about evaluating whether something is worth doing, but rather tuning 'not quite working' into 'working very well'. "

5-Axis Tuning:

Increase value(benefit items)
Reduce cost(cost items)
Change structure(model / path)
Extend time(cycle / term)
Change scope(scale / boundary)

4.4 7 Complete Cases (Candy)

Kong Yuan Medical Aesthetics Salary Cut: 5 Iterations
Chai Xiang Desert Study Tour: 5 Canvas Versions
Student Xu Wallet Partnership: ROI Is a Ruler, 50 Million Turnaround
Jia Hongyang Ten-Million-Level: 5 Rounds of Iteration
Yu Minmin Cross-City Branch: Role-Play Breakthrough + Employee ROI Reverse Trigger
Du Jinke Self-Service Chess & Card Room: Parallel A/B Testing + Exponential vs Linear Variables
Liang Jing Factory "5-Axis Tuning Canvas When ROI Doesn't Add Up"

5. Wang Kai RSI AI Decision Three Layers (Exclusive Complete Version)

Wang Kai RSI AI Investment Strategy and the "Good Asset, Good Price, Good Volatility" Framework §3:

5.1 Three-Layer Framework

好资产(前提)→ 好价格(安全边际)→ 好波动(入场信号)

Layer	Content
Good asset	Over 100 billion USD market cap + monopoly in a niche segment + long history + has experienced volatility proving long-term value
Good price	"Being cheap enough is not the key; understanding it well enough is the key" (Duan Yongping) / 6000-word intrinsic value Prompt
Good volatility	AI scores "risk" on a 1–5 scale / 5 risk variables

5.2 4 Uses of AI (Ranked by Importance)

As a ruler(most valuable) — quantifies fuzzy variables
Interrogate the decision-maker — "Why buy? What is the basis? What are the risks?"
Data analysis assistant(Claude Code / Cursor)
Execution assistant

5.3 5 Risk Quantification Variables (Built into the Prompt)

Whether the company's fundamentals have actually changed
Macro risk (Federal Reserve / tariffs / war)
Industry risk (replacement by AI / policy)
Liquidity risk
Position concentration

5.4 Capital Allocation Rules (Dynamic)

Risk level	Position
Risk 1	Buy full position
Risk 2	3/4
Risk 3	1/2
Risk 4	Reduce position
Risk 5	Clear position

6. How to Choose — ROI Paths by Identity

Who you are	Primary method	Starting point
Personal decisions	yitang list-deduce-build-check + 9 typical scenarios	3 personal-section questions (job change / school district house / MBA)
Team decisions	yitang triangle + Truman 4 layers of consensus	3 team-section questions (weekly meeting / KPI / lunch)
Company strategic decisions	yitang altitude section + 5-step training	3 company-section questions (IP / livestreaming / co-founder)
Investment decisions	Wang Kai RSI three layers + AI as ruler	Good asset + good price + good volatility + 5 variables
Want to train ROI ability	yitang 5-step training model + 7 cases	Awareness → write first draft → refine in detail → build consensus → conduct review
When ROI doesn't add up	Truman 5-axis tuning	Increase value + reduce cost + change structure + extend time + change scope

General discipline:

Decisions can be half a beat slow(Truman)
Absolute rational thinking by default(Wang Kai)
AI is a ruler, not a fortune teller(Wang Kai)
ROI is a ruler; tune 'not quite working' into 'working very well'(Truman)
4 progressive layers of consensus(action → judgment → logic → information)
If consensus is formed but ROI doesn't add up, abandon immediately(Wang Kai)
Write out the reasoning process(slow-thinking checkpoint)

7. Reference List

Main original author wiki:

Yitang / Complete Scientific Decision ROI System (4 main lessons + 5 Candy case sets) —— Truman 5-lesson series (cognition/breadth/depth/height/implementation training) + list-deduce-build-check + Type X vs Type Y + 5-step training model + 7 cases
Wang Kai Decision-Making System in the AI Era —— 5 core mindsets / standing from an AGI perspective / using AI as a measuring stick / absolute rationality / rules over outcomes / understanding = being able to do it
Wang Kai RSI AI Investment Strategy and the "Good Asset, Good Price, Good Volatility" Framework —— Three layers of RSI / 4 ways to use AI / 5 variables for risk quantification / evolution of the 250,000 USD capital pool
Wang Kai RSI Investment Real-World Case Collection + AI Value Investing Philosophy —— 6000+ word Prompt for intrinsic value calculation / IFS psychological support

History

2026-05-18:Topic-dimension wiki first release. Synthesized and rewritten from 4 author wikis. Explicitly marks 2 points of divergence: waiting target / adjusting and fixing when ROI calculation does not work vs giving up