关于「实战工具与检查器规格:鲸鱼显影台账、四件套与预注册证伪」,《鲸鱼客群与尊享档》给出的定量结论是什么?

结论:第八章给出蒸馏小工具「鲸鱼尾部检验与尊享档自选菜单试算器」、按四件套分组的检查器规格、每周鲸鱼显影台账与预注册检验,最贵的一项需要 14,579 位客户。高净值信号只用买家在本站自愿做出的行为,零敏感资料、零外购财富数据、零按受保护特征分流或差别定价。书中以标注框区分已发表的证据、可运行的 Python 计算(22 段程序)、原典与作者的假说。

数据口径:核对原始文献之后做了蓝图修正:幂律必须先检验,一本纯对数正态的合成账簿通过了幂律拟合优度检验,p 值 0.430,只有似然比检验给出 z 为 -1.271、无法裁决;前 2 % 的利润份额随尾部形状从 0.121 变到 0.768,只能按区间报;「自动显影」用 BG/NBD 模型认出下一年前 2 % 客户的命中率是 0.625,不是 1;诱饵效应在自然呈现下可能减弱或反转;「隐式探针」改写为对所有访客公开、可自选的入口。 书中参数为示意,全部计算由书中程序生成。

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: 14579

test 1, power 0.839,981 visitors per arm, 10.0 weeks at 4,000
test 1, power 0.955,380 visitors per arm, 13.8 weeks at 4,000
test 2, power 0.8138 follow-ups per group
test 2, power 0.9185 follow-ups per group
test 3, power 0.82,075 tail points, 14,579 customers
test 3, power 0.92,778 tail points, 19,518 customers
查看完整验算程序
# Chapter 8. p82: sample sizes for the pre-registered tests.
# Test 1: ordinary conversion is non-inferior after the VIP entry goes
# live: p = 0.030, margin 0.003, one-sided alpha 0.05;
# n per arm = (z_a + z_b)^2 * 2 p (1 - p) / m^2.
# Test 2: follow-up of request-form senders converts better than
# follow-up of entry clickers, 0.35 against 0.20, two-sided 0.05.
# Test 3: Vuong test of lognormal against power law: the per-point
# effect delta = z / sqrt(m) = -1.271 / sqrt(427) comes from book LN in
# p13; tail points needed = ((z_a + z_b) / delta)^2, customers needed =
# tail points / (427 / 3,000).
import math
from statistics import NormalDist

ND = NormalDist()


def n_noninf(p, m, alpha, power):
    za, zb = ND.inv_cdf(1 - alpha), ND.inv_cdf(power)
    return math.ceil((za + zb) ** 2 * 2 * p * (1 - p) / m ** 2)


def n_two(p1, p2, alpha, power):
    za, zb = ND.inv_cdf(1 - alpha / 2), ND.inv_cdf(power)
    pb = (p1 + p2) / 2
    a = za * math.sqrt(2 * pb * (1 - pb))
    b = zb * math.sqrt(p1 * (1 - p1) + p2 * (1 - p2))
    return math.ceil((a + b) ** 2 / (p1 - p2) ** 2)


if __name__ == "__main__":
    for pw in (0.8, 0.9):
        n = n_noninf(0.030, 0.003, 0.05, pw)
        print("test 1, power {}: {:,} visitors per arm, {:.1f} weeks at"
              " 4,000".format(pw, n, n / 4000))
    for pw in (0.8, 0.9):
        print("test 2, power {}: {:,} follow-ups per group".format(
            pw, n_two(0.35, 0.20, 0.05, pw)))
    delta = -1.271 / math.sqrt(427)
    for pw in (0.8, 0.9):
        za, zb = ND.inv_cdf(0.975), ND.inv_cdf(pw)
        m = math.ceil(((za + zb) / delta) ** 2)
        print("test 3, power {}: {:,} tail points, {:,} customers".format(
            pw, m, math.ceil(m / (427 / 3000))))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《帕累托鲸鱼客群自动显影、高净值信号自筛选与十倍客单尊享层跃迁母典》第 8 章「实战工具与检查器规格:鲸鱼显影台账、四件套与预注册证伪」(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-399 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026399,
  author    = {Fan, Yafu},
  title     = {Pareto Whale Auto-Detection, High-Net-Worth Self-Selection Signals and Ten-X VIP Tier Escalation},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论