关于「事后对上值多少证据:似然比、选择性记忆与巴纳姆句式」,《梅花易数的哈希与基率》给出的定量结论是什么?

结论:第六章用似然比为「事后对上」定价,并模拟选择性记忆如何把一半对一半的运气变成 0.795 的回忆相符率。

数据口径:书中的人群分布、注意率与记忆率均为示意;全部计算由书中程序生成;本书分析结构与概率,不做任何个人占断,也不构成商业、法律或理财建议。 第六章用似然比为「事后对上」定价,并模拟选择性记忆如何把一半对一半的运气变成 0.795 的回忆相符率。

English version

对照数据表

下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库与结论吻合的打印值: 0.795

true match rate0.500
remembered match rate, expected0.795
remembered match rate, simulated0.794
simulated 5th and 95th percentiles 0.727 0.853
clients who recall more than 0.75872 of 1,000
without reinterpretation, expected0.727
查看完整验算程序
# Chapter 6. p62: selective memory turns a coin into a reputation.
# Each of 1,000 simulated clients makes 200 readings; each reading
# matches later events with chance 0.5 (pure chance). A match is
# remembered with chance 0.8, a miss with chance 0.3, and a remembered
# miss is reinterpreted as a match with chance 0.25 ("it fitted after
# all"). All rates are illustrative. Fixed seed; exact expectation too.
import math
import random

rng = random.Random(20261112)
P_HIT, R_HIT, R_MISS, STRETCH = 0.5, 0.8, 0.3, 0.25


def main():
    rates = []
    for _ in range(1000):
        good = bad = 0
        for _ in range(200):
            if rng.random() < P_HIT:
                if rng.random() < R_HIT:
                    good += 1
            elif rng.random() < R_MISS:
                if rng.random() < STRETCH:
                    good += 1
                else:
                    bad += 1
        rates.append(good / (good + bad))
    rates.sort()
    mem_hit = P_HIT * R_HIT + (1 - P_HIT) * R_MISS * STRETCH
    mem_all = P_HIT * R_HIT + (1 - P_HIT) * R_MISS
    print("true match rate                     {:.3f}".format(P_HIT))
    print("remembered match rate, expected     {:.3f}".format(
        mem_hit / mem_all))
    print("remembered match rate, simulated    {:.3f}".format(
        math.fsum(rates) / len(rates)))
    print("simulated 5th and 95th percentiles  {:.3f}  {:.3f}".format(
        rates[49], rates[949]))
    above = len([r for r in rates if r > 0.75])
    print("clients who recall more than 0.75   {} of 1,000".format(above))
    no_stretch = P_HIT * R_HIT / mem_all
    print("without reinterpretation, expected  {:.3f}".format(no_stretch))


if __name__ == "__main__":
    main()

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《梅花易数体用生克自动机、年月日时哈希碰撞与外应注意捕获基率》第 6 章「事后对上值多少证据:似然比、选择性记忆与巴纳姆句式」(版本 1.0.0). 5D Library. DOI 待发布

DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 5DL-B-156 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026156,
  author    = {Fan, Yafu},
  title     = {Plum-Blossom Numerology as a Body-Use Automaton, Hash Collisions in Date-Time Casting and Base Rates of Attention-Captured Omens},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论