What does Birth-Time Error and Boundary Flips conclude, in numbers, about “Cell Widths and Stacked Layers: Ke, Double Hours and Rising Signs”?

Answer: Chapter 3 extends the grid to ke, clock hours and rising signs; with the layers stacked, 0.7230 of people have at least one layer that differs.

Scope of the numbers: Its parameters are illustrative, and all calculations are generated by its programs. Chapter 1 shows that in the illustrative model 0.2782 of remembered times fall exactly on a double-hour boundary, against 0.0083 without rounding.

中文版

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: 0.7230

share of people whose reported chart differs from the true one
double hour0.1736
clock hour0.3471
quarter hour0.7076
rising sign0.1259
any of the four0.7230
double hour or rising sign, if independent: 0.2776
day-pillar convention split (23:00 vs 00:00): 0.0433

Key formula

\[P_{\text{flip}} = 1 - \frac{\sigma}{w}\left[x\,(2\Phi(x)-1) + 2\varphi(x) - 2\varphi(0)\right],\qquad x = w/\sigma .\]
Show the full program
# Chapter 3. p33: a chart has several layers, and any one can flip.
# Fixed-seed Monte Carlo of the memory source of p11 (illustrative
# weights, recall error 15 minutes). For each simulated person we
# compare the cell of the true time and of the report in four layers:
# double hour (boundaries at odd hours), clock hour, quarter hour, and
# the rising sign at latitude 31.2 (sign lengths from p32; the sky's
# phase at midnight is drawn at random, so its boundaries are not tied
# to the clock). We also count a pure convention split: the day pillar
# changes at 23:00 in one school and at 00:00 in another, so births in
# 23:00-24:00 get different day pillars with no error at all.
import bisect
import math
import random

GRAN = [1, 5, 15, 30, 60]
W, S = [0.10, 0.15, 0.10, 0.25, 0.40], 15.0
# 纬度 31.2 度的十二宫上升时长(时钟分钟,取自 p32)
SIGN_MIN = [83.2, 96.2, 118.9, 137.9, 142.4, 139.5, 139.5, 142.4, 137.9,
            118.9, 96.2, 83.2]
DAY = math.fsum(SIGN_MIN)
EDGES = [math.fsum(SIGN_MIN[:k + 1]) for k in range(12)]


def sign_at(t, phase):
    x = (t * DAY / 1440.0 + phase) % DAY
    return bisect.bisect_right(EDGES, x)


rng = random.Random(20261113)
N = 100000
hits = {"double hour": 0, "clock hour": 0, "quarter hour": 0,
        "rising sign": 0, "any of the four": 0}
conv = 0
for _ in range(N):
    t = rng.random() * 1440.0
    g = rng.choices(GRAN, W)[0]
    r = (math.floor((t + rng.gauss(0.0, S)) / g + 0.5) * g) % 1440
    phase = rng.random() * DAY
    f = [int((t + 60) % 1440 // 120) != int((r + 60) % 1440 // 120),
         int(t // 60) != int(r // 60),
         int(t // 15) != int(r // 15),
         sign_at(t, phase) != sign_at(r, phase)]
    for name, x in zip(("double hour", "clock hour", "quarter hour",
                        "rising sign"), f):
        hits[name] += x  # step
    hits["any of the four"] += any(f)  # step
    conv += t >= 1380.0  # step
print("share of people whose reported chart differs from the true one")
for name in ("double hour", "clock hour", "quarter hour", "rising sign",
             "any of the four"):
    print("  %-17s %.4f" % (name, hits[name] / N))
print("double hour or rising sign, if independent: %.4f"
      % (1.0 - (1.0 - hits["double hour"] / N)
         * (1.0 - hits["rising sign"] / N)))
print("day-pillar convention split (23:00 vs 00:00): %.4f" % (conv / N))

Single source and citation

Fan, Y. (2026). Error Propagation from Remembered Birth Times and Robust Chart Fixing: Flip Probabilities at Double-Hour Boundaries and the Fewest Past-Event Questions (Version 1.0.0), Chapter 3: Cell Widths and Stacked Layers: Ke, Double Hours and Rising Signs. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-284 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026284,
  author    = {Fan, Yafu},
  title     = {Error Propagation from Remembered Birth Times and Robust Chart Fixing: Flip Probabilities at Double-Hour Boundaries and the Fewest Past-Event Questions},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book