5D Library › Power-Law Weighting and Bottom-Up Controllability in Multi-Tier Conditional Excitation Systems · 中文

Preface: Five Tiers, Two Directions

Note

Please read this first. This book presents mathematical models for teaching and research. It is not investment, trading, career, tax or legal advice, and it must not be used to select, screen or rate any particular person, team or start-up. Except where a source is cited, the “people”, “tier scores”, “coupling coefficients” and data in this book are invented or generated by the programs, and the parameters are illustrative; simulated and back-tested results do not represent the future and promise no success. The book offers no method that “guarantees” anything, “always wins” or is “proven optimal”: every mathematical result holds only under the assumptions stated with it.

Chapter 3 discusses polygenic scores for educational attainment. Such scores are population-level statistical tools: they predict poorly for individuals and transfer poorly between ancestry groups, and they must not be used for admissions, hiring, partner choice or any other decision about an individual. The book argues neither that “nature decides everything” nor that “effort can do anything”: the first conflicts with the within-family and environmental evidence, the second with the mathematics of controllability. If you have been distressed for a long time about your prospects, about unemployment or about having “bad fate”, please talk to someone you trust or to a professional.

Many people carry a pyramid in their heads. At the top sit “talent” and “background”; below them “timing”, “environment” and “connections and reputation”; at the bottom, the one thing under one’s own control, “learning and effort”. The author calls it the “bidirectional pyramid of the five success factors”: read downwards, the higher a tier, the more it weighs in the outcome; read upwards, investment at the bottom can push through the tiers one by one and finally “unlock” the top.

This book turns the pyramid into mathematical models that can be computed and tested. It asks four questions. First, how do five conditions combine into an outcome: do they add or multiply? Second, can the “weight of each tier” be defined and estimated, and what do the author’s figures of 50 %, 25 %, 12.5 %, 7.5 % and 5 % mean? Third, what does published research say about endowment, timing, environment, reputation and effort? Fourth, in a control system, which tiers can an input applied at the bottom reach, how far can it move them, and how much “energy” does it take?

What this book covers

The five tiers are written C1C_1 endogenous prior, C2C_2 timing and cycles, C3C_3 environmental topology, C4C_4 reputational games and C5C_5 knowledge investment. The five-tier structure, the way the tiers are coupled, the weights and the “bidirectional” claim are all the author’s conceptual model; throughout the book they are labelled as the author’s model or hypotheses, and the evidence sections use published research only.

Chapter 1 writes success as a product of five factors and as a chain of conditional probabilities, derives the weakest-link property and log-additivity of the multiplicative model and compares it with an additive one; in 1957 Shockley explained the log-normal distribution of research output by multiplying several factors. Chapter 2 examines the author’s “binary” weights and uses Shapley values (LMG) and Sobol’ indices to show that variance shares depend on the correlations, on the order of entry and on the scale. Chapter 3 reviews the evidence on endowment and timing: the explanatory power of polygenic scores for educational attainment and how it shrinks within families, and the long-run earnings losses of graduating in a recession. Chapter 4 covers environment, reputation and luck: the Moving to Opportunity experiment, the Matthew effect and the artificial music market, and the “talent versus luck” model. Chapter 5 uses the linear control system ẋ=Ax+Bu\dot{x} = Ax + Bu to study bottom-up controllability: Kalman’s rank condition, the controllability Gramian and control energy. Chapter 6 separates “potential” from “realization” and shows that different coupling assumptions lead to opposite conclusions. Chapter 7 reads these discussions beside Wang Chong’s Lunheng (chapter “Ming lu”), Yuan Liaofan’s Liaofan si xun (“Liming zhi xue”) and the Xunzi (“Quan xue”). Chapter 8 gives a four-step checklist and a pre-registered randomized controlled trial.

The book borders on several other volumes of 5D Library; where they overlap it only summarizes and cites them, without repeating their programs. Volume 13, Classical and Statistical Mechanics of Human Action, discussed the dynamics of action; Volume 15, The Sandbox, Determinism and Active Inference, discussed pessimistic priors and the sense of fate; Volume 23, The 85% Rule for Optimal Learning, Myelination Kinetics and Deliberate Practice, discussed deliberate practice (Chapter 4 cites only its conclusions); Volume 25, Social Network Topology, Structural Holes and Complex Contagion, discussed network position (Chapter 4 only reviews it for the environment tier); Volume 35, Evolutionary Game Theory, Direct and Indirect Reciprocity and the Emergence of Cooperation, discussed reputation and indirect reciprocity; Volume 37, Two-Scale Renormalization, Occupancy Entropy and Thrust-Wave Superposition in Discrete Cyclical Networks, discussed cycles; and Volume 33, Scientific Methodology, Falsifiability and the Mathematical Statistics of the Replication Crisis, is the methodological charter of the series, whose way of writing decision rules Chapter 8 follows.

Claims corrected after checking

Several points in the author’s framework needed to be corrected or qualified after checking. Each chapter says so where it arises, and the list is collected in the editorial notes published with the book.

  1. The “binary power-law” weights. A strict binary (halving) sequence is 50 %, 25 %, 12.5 %, 6.25 %, 3.125 %, which sums to 96.875 % and leaves 3.125 % unassigned; the author’s 7.5 % and 5 % are not binary but an approximation that makes the total 100 %. Moreover, 2−k2^{-k} is a geometric (exponential) sequence, not a power law k−αk^{-\alpha} in the usual sense, and the two differ markedly in the fourth and fifth tiers. The weights are the author’s assumption, not an empirical estimate from any study. The book keeps the author’s wording in the title, explains the difference in the text, and asks the author in the editorial notes to confirm which set of numbers to use (Chapter 2).
  2. Variance shares are not fixed. Once the tiers are correlated, “how much variance each tier explains” depends on the method of splitting: in the illustrative example of Chapter 2 the same model gives C5C_5 2.0 % when split top-down, 20.6 % when split bottom-up and 10.9 % by the Shapley value (LMG). On the raw scale the multiplicative model also has interaction terms, so the answer depends on the scale as well. “Upper tiers matter more” therefore cannot be derived from the mathematics of the weights; it can only be estimated from data, and the estimate depends on the form of the model (Chapter 2).
  3. “Five-tier tensor product.” Multiplying five scalars is an ordinary product, not a tensor product in the mathematical sense; and the chain P(S)=P(C1)P(C2∣C1)⋯P(S) = P(C_1)P(C_2 \mid C_1)\cdots is an identity that holds in any order, so it implies no top-down hierarchy by itself (Chapter 1).
  4. “Prove how bottom-tier knowledge investment C5=100%C_5 = 100\% unlocks the top-tier prior C1C_1 in reverse.” This cannot be proved, and under the author’s own assumptions its opposite can: if C1C_1 is an unchangeable endowment, it is an uncontrollable state of the linear model, the Gramian is zero in that direction, and no input can move it. What can be raised is the realization of C1C_1 (the product of the potential with the other tiers), and realization has a ceiling: it cannot exceed the potential itself. Controllability depends only on the coupling structure. Only if the prior is assumed to be plastic does C1C_1 become controllable, and then the energy required is thousands of times that needed to move C5C_5; under a “crowding-out” assumption, more knowledge investment even lowers the overall outcome (Chapters 5 and 6).
  5. The numbers in the evidence. Lee and colleagues (2018) report that a polygenic score for educational attainment explains 11–13 % of the variance, and Okbay and colleagues (2022) report 12–16 %; both figures check out. But they are population-level figures: Okbay and colleagues found that the direct effect within families is only about half the magnitude of the population association; Kong and colleagues found that alleles parents did not transmit to their children are also associated with the children’s attainment, at about 29.9 % of the transmitted effect (“genetic nurture”); and the scores predict markedly less well in samples of non-European ancestry. The figures are therefore neither the whole of the “endowment tier” nor endowment alone (Chapter 3).
  6. Timing has large effects but small variance shares. The long-run earnings losses of graduating in a recession (Kahn 2010; Oreopoulos, von Wachter and Heisz 2012) are large for the cohorts affected, but the illustrative calculation of Chapter 3 suggests that their share of the variance of earnings across everyone may be below 1–2 %. That supports neither “timing is 25 %” nor “timing does not matter”: a variance share is not a measure of importance to the individual.
  7. “Talent versus luck” is a model, not evidence. The conclusions of Pluchino and colleagues (2018) follow from the rules they assume; the program in this book is a mean-field simplification of their model and serves only to show what distributions multiplicative good and bad luck produce (Chapter 4).
  8. Classical sources. “If one’s fate is to be poor and lowly, then even if made rich and noble one will still run into calamity” comes from the “Ming lu” chapter of the Lunheng, not from “Qi shou”, which discusses endowed qi and length of life. “Fate is made by oneself; fortune is sought by oneself” is, in “Liming zhi xue”, spoken to Yuan Liaofan by the Chan master Yungu, not coined by Yuan himself, and the collective title Liaofan si xun (Liaofan’s Four Lessons) was given only when the essays were later printed together (Chapter 7).

How to read the boxes

As in the other books of 5D Library, different kinds of statement are marked separately.

Evidence

Evidence boxes report published research findings and cite their sources.

Calculation

Calculation boxes give formulas and short Python programs. Every program runs as printed in a standard Python 3 environment, uses only the standard library, and is followed by its actual output. The “people” and data in the book are invented or generated by the programs.

Hypothesis · not established

Hypothesis boxes contain the author’s ideas, are marked “not established”, and state what result would show them to be wrong.

Source text

Source boxes quote classical texts with the author’s plain-language explanation.

Note

Note boxes give cautions and explain choices.

Where the book applies concepts from control theory, statistical mechanics and network science (coupling, controllability, energy, cascades) to individuals and societies, they are models and analogies. They help to state intuitions clearly and to separate out what can be tested, but a life is not a linear system and the five tiers are not physical quantities that can be measured directly. The mathematical properties of a model are not predictive evidence about reality.

About this edition

This is version 1.0.0, the first complete edition: a preface and eight chapters, in Chinese and English. The text, boxes and code were drafted with an AI writing tool from the author’s title, framework and notes; the author is responsible for reviewing and revising them and for checking every source. Corrections and missing references are welcome through the 5D Library website.

Disclosure

The author uses the works of 5D Library, including this book, as the methodological basis of his community courses, the 5D Ascension Cards (Dimensional Shift Card Decks) and the Yafu AI knowledge base, and therefore has a commercial interest in the subject of this book. The author also uses this “bidirectional pyramid” framework in his community courses and consulting, and founded and runs Asia Influencer X and Yafu Global (discovery and management of influencers and talent), and therefore has a direct commercial interest in the subject of this book. The book promotes no course, card deck or service; it uses no data from the author’s students or clients; the Chapter 8 trial is designed to be run and analysed by researchers independent of the author’s business and does not recruit from the author’s community, students or clients (including the clients and signed talent of these businesses); all the code and the study protocol are published openly for anyone to check. All models in this book are offered solely for mathematical and statistical discussion; they constitute no career, educational, investment or talent-management advice, and the book makes no promise about the outcome of any person or project.

Fan, Y. (2026). Power-Law Weighting and Bottom-Up Controllability in Multi-Tier Conditional Excitation Systems: A Mathematical Synthesis of Endogenous Priors, Timing Windows, Network Topology and Reputational Games [多层条件激发系统中的幂律权重与逆向控制可及性] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23810093 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).