5D Library › Retrodictive Belief Locking, Sequential Pain-Point Agitation and Markovian Upsell Automata · 中文
Preface: Where Trust Comes From, and Where Upsells Lead
Note
This is not divination, nor marketing or business advice; please read this first. The book studies three mechanisms of reading services that are often treated as “conversion techniques”: winning trust with statements about the client’s past, pushing upsells with price steps and sunk cost, and the conversion chain from a single reading to a course to a business licence. Every psychological parameter, conversion rate and price in it is a model parameter, not data from any business; all data are generated by the programs printed in the book. The book neither claims nor tests that any reading comes true, and it computes the costs of each mechanism – to users and to the business – together with its gains.
A sceptical user pays for a first reading. The reader names the hardest of her last fifteen years, and a few old events besides; she begins to believe. Then come a deep report, a course, perhaps in the end a business licence. The author’s outline writes this path as three models: a Bayesian “trust avalanche”, a commitment ratchet, and a three-stage Markov conversion chain. This book writes the three as runnable programs and asks at each step: is the user’s trust evidence? How much of an upsell’s profit is left after refunds, churn and harm? Do the chain’s predictions survive contact with data?
Like the other books of 5D Library, it tries to refute first, then sets bounds, and reports negative results as they are.
What the book contains
Chapter 1 computes the likelihood ratio of a retrodictive statement and contrasts the “trust avalanche” under true and under felt likelihood ratios. Chapter 2 gives a fair sequential test of retrodictive ability and computes the “accuracy” that selective recall produces. Chapter 3 sets up an upsell model with a resistance threshold and sunk cost and finds the best price ratio and timing. Chapter 4 computes what “pain-point agitation” buys and costs, and two guardrails: a cooling-off period and a distress screen. Chapter 5 writes “reading, course, licence” as an absorbing Markov chain and computes conversion probabilities and lifetime value. Chapter 6 deals with estimating transition matrices, the course’s cannibalization of readings, and heterogeneous clients. Chapter 7 reads the “Mo” and “Fei qian” chapters of the Guiguzi and the “Shuo nan” chapter of the Han Feizi. Chapter 8 gives a calculator a user can run herself and a pre-registered blinded test.
Claims corrected on checking
- “Consecutive retrodictive hits collapse a sceptic’s defences in a phase transition” is a transition in feeling, not in evidence. A generic (Barnum) retrodictive statement has a true likelihood ratio of only 1.19, and a sceptic would need 27 hits to be half convinced; if it “feels accurate” (its base rate of 0.80 taken for 0.20), 3 suffice (p01). Among 20,000 simulated users, a reader with no insight at all convinces 0.5263 of them within 6 generic statements, while users who update with the true ratios are almost never convinced (p02).
- “The trough year of the past 15” looks specific but is not. If one year is the worst for a quarter of the population, a hit on it has a true likelihood ratio of 3.60, not the apparent 13.50 (p01).
- Selective recall turns chance into “accuracy”. Fifteen year-by-year statements made with no insight, right half the time, are remembered under ordinary selective recall as 0.7493 accurate; written down beforehand, accuracy is back at 0.4995 (p04).
- The best upsell price ratio lies far below the user’s resistance threshold, and lower still once refunds count. Sunk cost raises the tolerated ratio along the ladder (p05); timing matters as much as price (p06).
- Pain-point agitation looks profitable judged on the first sale; with refunds and churn counted its best intensity falls, but not to zero. The stronger the agitation, the larger the share of buyers in acute distress (p07); a cooling-off period and a distress screen together bring that share down to 0.056 (p08).
- In the chain “reading, course, licence”, the licence step carries most of a reader’s lifetime value. A reader becomes a licensee with probability only 0.0267 yet that path brings about three quarters of her value (p09); pooling different kinds of client into one chain badly underestimates long-tail retention (p13).
How to read the boxes
The boxes separate published evidence, runnable Python calculations (16 programs, standard library only, fixed random seeds, each run at build time with its real output printed beneath it), classical sources, notes and the author’s hypotheses; every hypothesis states what result would show it wrong. So that code stays intact in print, programs and their output use ASCII characters only: states are V (visitor), T (reader), C (student), L (licensee) and X (lapsed), and modules C, A, M and B.
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 develops and runs paid online reading software and services on traditional divinatory themes, an academy of chart-reading courses and business-practitioner licences; the conversion models of this book apply directly to those products, so he has a direct commercial interest in its subject. The book promotes no software, course, card deck or service, uses no user or sales data from the author’s businesses, and accepts no sponsorship. All its models are offered for mathematical study and teaching only and constitute no promise that any reading comes true and no marketing, pricing or other business advice; the guardrails for users in distress are the author’s commitment for his own business, not an option. The test in Chapter 8 must be carried out with pre-registered code by an independent analyst who has no financial relationship with the author and does not know where the statements came from, and its results will be published whether positive or negative.
Fan, Y. (2026). Retrodictive Belief Locking, Sequential Pain-Point Agitation and Markovian Upsell Automata: Bayesian Trust Avalanches, Sunk-Cost Commitment Ratchets and Cross-Domain Transition Matrices [回溯信念锁定、序贯痛点唤醒与马尔可夫升单自动机] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.24090186 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).