What does One-Site Thousand, Five-Site Matrix conclude, in numbers, about “A Good Week Is Not a Stable Thousand: Daily Noise, Sequential Tests and the Clone Gate”?
Answer: Under the illustrative parameters, one site's daily revenue has a coefficient of variation of 0.604, so a single day says almost nothing, and a 7-day mean gives a site that truly earns RM 850 a 0.218 chance of showing a thousand. A sequential test written for the negative binomial settles most cases in a little over 20 days with error rates as designed; wrongly assuming Poisson orders pushes the false-pick rate to 0.208. Of the four gates, the sequential test plus 28 days plus a floor of RM 950, with a false-pick rate of 0.090 and a recognition rate of 0.942, is the one this book recommends.
Scope of the numbers: Order values, conversion rates, fees, incident rates and market parameters are illustrative, and all calculations are generated by the book's programs. Revenue figures are illustrative arithmetic, not a promise of income, and not investment, legal or tax advice.
Data table
The numbers below are printed verbatim by the chapter's Python program and can be reproduced by running it. standard-library Python onlyprinted values that match the answer: 0.218, 0.604, 28, 850
| true mean RM 1,000: simulated daily mean RM 1,005.2, sd RM 607.6, cv 0.604 | ||||
| share of days below RM 700: 0.354; above RM 1,300: 0.265 | ||||
| window | sd of mean | P(RM 850 site shows >= 1,000) | P(1,000 site < 900) | |
| 3 d | RM | 346.9 | 0.284 | 0.428 |
| 7 d | RM | 231.3 | 0.218 | 0.347 |
| 14 d | RM | 157.8 | 0.147 | 0.281 |
| 28 d | RM | 113.7 | 0.070 | 0.191 |
Show the full program
# Chapter 2. p21: how noisy one site's daily revenue is around RM 1,000.
# Daily orders are gamma-mixed Poisson (negative binomial) with mean m and
# shape k = 4, so the variance is m + m * m / k; order values are lognormal
# with mean RM 89 and coefficient of variation 0.5. k, the value spread and
# the true means below are illustrative, not measured on any real site.
import math
import random
import statistics
RNG = random.Random(357021)
K, AOV, CV = 4.0, 89.0, 0.5
SIG = math.sqrt(math.log(1 + CV * CV))
MU = math.log(AOV) - SIG * SIG / 2
def poisson(lam):
limit, k, p = math.exp(-lam), 0, 1.0
while True:
p *= RNG.random()
if p <= limit:
return k
k += 1
def one_day(m):
n = poisson(RNG.gammavariate(K, m / K))
return math.fsum(RNG.lognormvariate(MU, SIG) for _ in range(n))
def window_means(true_rev, days, reps):
m = true_rev / AOV
return [statistics.fmean(one_day(m) for _ in range(days))
for _ in range(reps)]
days = [one_day(1000.0 / AOV) for _ in range(20000)]
mean, sd = statistics.fmean(days), statistics.pstdev(days)
print("true mean RM 1,000: simulated daily mean RM {:,.1f}, sd RM {:.1f}, cv {:.3f}"
.format(mean, sd, sd / mean))
print(" share of days below RM 700: %.3f; above RM 1,300: %.3f" % (
len([d for d in days if d < 700]) / len(days),
len([d for d in days if d > 1300]) / len(days)))
print("window sd of mean P(RM 850 site shows >= 1,000) P(1,000 site < 900)")
for w in (3, 7, 14, 28):
low = window_means(850.0, w, 4000)
high = window_means(1000.0, w, 4000)
print(" %2d d RM %6.1f %.3f %.3f" % (
w, statistics.pstdev(high), len([x for x in low if x >= 1000]) / 4000,
len([x for x in high if x < 900]) / 4000))Single source and citation
Fan, Y. (2026). The Thousand-Ringgit Single-Site Minimum Loop, Headless Core Multi-Skin Cloning and Five-Site Revenue Multiplication (Version 1.0.0), Chapter 2: A Good Week Is Not a Stable Thousand: Daily Noise, Sequential Tests and the Clone Gate. 5D Library. DOI pending
DOI pending · monograph page goes live after publication · 5DL-B-357 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026357,
author = {Fan, Yafu},
title = {The Thousand-Ringgit Single-Site Minimum Loop, Headless Core Multi-Skin Cloning and Five-Site Revenue Multiplication},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}Other chapter answers in this book
- Chapter 1 · Ask What a Thousand a Day Leaves: The Minimum Loop, Unit Contribution and Break-Even
- Chapter 3 · A Central Core Is Not Always Worth Building First: Headless Architecture, Information Hiding and Economies of Scope
- Chapter 4 · A New Skin Is Not a Copy of the Revenue: Per-Skin Accounts by Language, Country and Niche
- Chapter 5 · Five Sites That Do Not Raid Each Other: Orthogonal Partition of Keywords and Audiences, and the Multiplier
- Chapter 6 · The Price of a Shared Core: Correlated Failure, Blast Radius and the Chargeback Pool
- Chapter 7 · East-West Classical Sources: Sunzi on Division and Variation, and Smith on the Extent of the Market
- Chapter 8 · System Architecture and Operations Spec: The 72-Hour Clone Pipeline, the Weekly Ledger and the Five-Site Matrix Cloning Calculator