5D Library › Modular Numerology, Character-Stroke Hashes and Residue-Class Distributions · 中文
Chapter 8. A Four-Step Check and a Pre-Registered Matching Test
Four questions before trusting a numerology program
This chapter turns the book into four questions that can be put to any program for life path numbers, names or phone numbers, the author’s own included. They are not advice on any decision.
1. Which method does it use? Whether master numbers are kept, and whether year, month and day are reduced first or added all at once (Chapter 1); simplified or Kangxi strokes (Chapters 4 and 5); how 0 and 5 are treated (Chapter 6). These choices give the same input different numbers. A program should state them and be checkable against known examples.
2. What is the base rate of each label? How large a share of people is a “number 2 person” or holds a “good number” (Chapters 2, 5 and 6)? Only with the base rate can one judge how much it means that a statement fits a person (Kahneman and Tversky 1973).
3. How sensitive are the labels to convention? Under another stroke convention the total grid’s label changes for almost half of all names; adding one to a phone number changes its modulo-80 label more often than not (Chapters 5 and 6). A program should report how its results change under another reasonable convention.
4. Whom does it describe? Volume 63 showed that being judged accurate is not the same as telling people apart. In 1949 Forer had students rate a “personal” description that was the same for everyone, and the average accuracy was still high (Forer 1949). Whether a reading belongs to its owner must be measured by a blind matching test, not by accuracy ratings (Carlson 1985; Dean and Kelly 2003).
A pre-registered matching test
Hypothesis · not established
Study protocol. One method for life path numbers (one of A, B, C or D of Chapter 1) and one set of descriptions are registered publicly before recruitment and not changed afterwards. Each participant gives a date of birth; the researcher removes the numbers from the descriptions of the life path numbers, shuffles them, and asks the participant to pick the one most like themselves. Primary outcome: the share of participants who pick the description of their own number. Test: one-sided exact binomial test, , against the chance level. Decision: only reaching the critical number of hits counts as support for “the descriptions carry information about their owners”. The test is carried out by an independent analyst who has no financial relationship with the author and does not know the participants’ numbers, and the results are published whether positive or negative.
The chance level is where this test most easily goes wrong. Let be the share of people whose life path is and the share of participants who pick description ; if the descriptions carry no information, the hit rate is . If is even, the sum is one ninth whatever descriptions people prefer; if is uneven and people prefer some descriptions (the more flattering ones, say), the chance level is no longer one ninth.
Calculation
# Chapter 8. A pre-registered matching test, and its base-rate trap.
# Each participant reads the standard descriptions of the life path
# numbers, unlabelled and in random order, and picks the one most like
# themselves. A hit is a pick of the participant's own number. The
# chance hit rate is sum_k p_k q_k, where p is the population share of
# number k and q the share of participants who pick description k. If
# p is even, this is 1/9 whatever q is; if p is uneven (master numbers,
# rule A) and q favours some descriptions, chance is no longer 1/9.
# q below is an invented example of a preference for flattering texts.
import datetime as dt
from math import comb
MASTERS = (11, 22, 33)
KEYS = list(range(1, 10)) + [11, 22, 33]
Q = dict(zip(KEYS, (0.14, 0.03, 0.10, 0.04, 0.10, 0.05, 0.10, 0.10,
0.08, 0.16, 0.06, 0.04)))
def digit_sum(n):
return sum(int(c) for c in str(n))
def rule_a(y, m, d):
n = digit_sum(y) + digit_sum(m) + digit_sum(d)
while n > 9 and n not in MASTERS:
n = digit_sum(n)
return n
def shares(y0=1950, y1=2009):
c, n = {k: 0 for k in KEYS}, 0
day, end = dt.date(y0, 1, 1), dt.date(y1, 12, 31)
while day <= end:
c[rule_a(day.year, day.month, day.day)] += 1
n += 1
day += dt.timedelta(days=1)
return {k: v / n for k, v in c.items()}
def tail(n, k, p):
return sum(comb(n, i) * p ** i * (1 - p) ** (n - i)
for i in range(k, n + 1))
def plan(p0, p1, alpha=0.05, power=0.90):
for n in range(20, 5000):
k = next(k for k in range(n + 1) if tail(n, k, p0) <= alpha)
if tail(n, k, p1) >= power:
return n, k
if __name__ == "__main__":
p = shares()
chance_a = sum(p[k] * Q[k] for k in KEYS)
print(" chance hit rate, nine descriptions, even base rates: {:.4f}"
.format(1 / 9))
print(" chance hit rate, twelve descriptions under rule A,")
print(" with the example preferences q: {:.4f}".format(chance_a))
print(" (twelve descriptions and even preferences would give {:.4f})"
.format(sum(p[k] for k in KEYS) / 12))
print(" example preferences q (numbers 1-9, 11, 22, 33):")
cells = ["{}:{:.2f}".format(k, Q[k]) for k in KEYS]
print(" " + " ".join(cells[:6]))
print(" " + " ".join(cells[6:]))
print(" sum of q = {:.2f}".format(sum(Q.values())))
n, k = plan(1 / 9, 0.16)
print(" nine descriptions, hit rate 0.16 against 1/9, one-sided"
" 0.05,")
print(" power 0.90: {} participants, at least {} hits".format(n, k))
print("\nReading: with plain residues the chance level is exactly 1/9")
print("and a pre-registered test needs a few hundred people. With")
print("master numbers the chance level depends on which descriptions")
print("people like, so it must be estimated by shuffling, not assumed.")
# chance hit rate, nine descriptions, even base rates: 0.1111
# chance hit rate, twelve descriptions under rule A,
# with the example preferences q: 0.0945
# (twelve descriptions and even preferences would give 0.0833)
# example preferences q (numbers 1-9, 11, 22, 33):
# 1:0.14 2:0.03 3:0.10 4:0.04 5:0.10 6:0.05
# 7:0.10 8:0.10 9:0.08 11:0.16 22:0.06 33:0.04
# sum of q = 1.00
# nine descriptions, hit rate 0.16 against 1/9, one-sided 0.05,
# power 0.90: 419 participants, at least 58 hits
#
# Reading: with plain residues the chance level is exactly 1/9
# and a pre-registered test needs a few hundred people. With
# master numbers the chance level depends on which descriptions
# people like, so it must be estimated by shuffling, not assumed.With the plain digital root (method D) and nine descriptions, the chance hit rate is exactly 0.1111, whatever the participants’ preferences. With method A, master numbers kept, and twelve descriptions, the base rates are uneven; under the preferences used as an example in the program (the twelve are printed, summing to 1.00), the chance hit rate is 0.0945, not the one twelfth of 0.0833. Taking 0.0833 as the chance level by mistake would make a set of descriptions with no information at all look “better than chance”. So the protocol should register the plain digital root, or estimate the chance level by a permutation test that shuffles participants against dates, rather than assume it.
With the plain digital root, nine descriptions, one-sided 0.05 and power 0.90, detecting a hit rate of 0.16 against one ninth needs 419 participants and at least 58 hits.
Note
What the test tests, and what it does not. It tests whether the description of a person’s life path number is more like that person than other people’s descriptions are. It does not test whether life path numbers predict any event. A negative result would not show that what users feel when reading the descriptions is unreal; it would show only that the feeling does not depend on whether the description is their own.
Summary
Before trusting a numerology program, ask four questions: which method it uses, what the base rates of its labels are, how sensitive the labels are to convention, and whom it describes. A pre-registered matching test can answer the last question; its chance level is one ninth when base rates are even and must be estimated by permutation when master numbers are kept. Detecting a hit rate of 0.16 at one-sided 0.05 with power 0.90 takes 419 participants.
References
Fan, Y. (2026). Modular Numerology, Character-Stroke Hashes and Residue-Class Distributions: Digital-Root Arithmetic, Kangxi Radical Restorations, Modulo-80 Grid Mappings and Empirical Base-Rate Concentration [姓名笔画、生命灵数与手机号码的同余映射和分布论] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.24090178 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).