关于「不等概率的收集:容斥公式、泊松化积分与款式倾斜」,《盲盒收集与福袋概率公示》给出的定量结论是什么?
结论:第三章用容斥公式与泊松化积分算不等概率,12 款加 1/144 隐藏款的全套期望是 149.07 盒;
数据口径:书中参数为示意;全部计算由书中程序生成;本书是非个人化法律或理财建议,法条以上架前人工复核为准。 第一章用调和数、精确分布与固定种子模拟算等概率集齐;第二章用几何分布算隐藏款等待,并陈复合事件判断偏差的相反证据,说明保底只改变等待(保底 144 次时平均 91.21 次)而不改变标称概率;第三章用容斥公式与泊松化积分算不等概率,12 款加 1/144 隐藏款的全套期望是 149.07 盒;第四章用双杯问题算合买与重复款,以及开盒止损;第五章算福袋的期望、方差与低于售价的概率(示意福袋为 0.8280)以及小概率加权。
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: 12, 144, 149.07
| rates: hidden 1/144, each regular 143/1728 | 1/12.084 |
| E[boxes] for the 12 regulars only | 37.4989 |
| (12 * H_12 | 37.2385 scaled by 144/143 = 37.4989) |
| E[boxes] for all 13 incl. hidden | 149.0732 |
| E[wait for hidden alone] | 144.0000 |
| extra over the hidden alone | 5.0732 |
| hidden rate E[regular set] E[full set] full/hidden-alone | |
| 1/24 | 38.86 44.50 1.8540 |
| 1/48 | 38.03 60.77 1.2661 |
| 1/72 | 37.76 81.26 1.1287 |
| 1/144 | 37.50 149.07 1.0352 |
| 1/288 | 37.37 290.66 1.0092 |
查看完整验算程序
# Chapter 3. p31: unequal probabilities by inclusion-exclusion.
# For designs with probabilities p_1..p_m (summing to 1) the expected
# number of boxes to see every design is
# E[T] = sum over nonempty S of (-1)^(|S|+1) / P(S),
# with P(S) the total probability of the designs in S (Flajolet, Gardy
# and Thimonier 1992). Exact with Fractions; subsets are grouped by
# how many regular designs they contain, so 13 designs need 26 terms.
# Series: 12 regulars sharing 143/144, one hidden at 1/144 (illustrative).
import math
from fractions import Fraction
H = Fraction(1, 144)
R = (1 - H) / 12
def expected(n_reg, r, hidden):
total = Fraction(0)
for j in range(n_reg + 1):
for h in (0, 1) if hidden else (0,):
size = j + h
if size == 0:
continue
ps = j * r + (h * hidden if h else 0)
term = Fraction(math.comb(n_reg, j)) / ps
total += term if size % 2 == 1 else -term
return total
e_reg = expected(12, R, None)
e_all = expected(12, R, H)
print("rates: hidden 1/144, each regular {} = 1/{:.3f}".format(
R, float(1 / R)))
print("E[boxes] for the 12 regulars only: {:.4f}".format(float(e_reg)))
print(" (12 * H_12 = 37.2385 scaled by 144/143 = {:.4f})".format(
37.2385 * 144 / 143))
print("E[boxes] for all 13 incl. hidden: {:.4f}".format(float(e_all)))
print("E[wait for hidden alone]: {:.4f}".format(float(1 / H)))
print("extra over the hidden alone: {:.4f}".format(
float(e_all - 1 / H)))
print()
print("hidden rate E[regular set] E[full set] full/hidden-alone")
for n in (24, 48, 72, 144, 288):
h = Fraction(1, n)
r = (1 - h) / 12
a = expected(12, r, None)
b = expected(12, r, h)
print(" 1/{:<4d} {:8.2f} {:8.2f} {:.4f}".format(
n, float(a), float(b), float(b / n)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《盲盒收集券问题、隐藏款几何等待时间与福袋概率公示合规母典》第 3 章「不等概率的收集:容斥公式、泊松化积分与款式倾斜」(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-210 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026210,
author = {Fan, Yafu},
title = {Blind-Box Coupon Collecting, Geometric Waits for Hidden Variants and Odds Disclosure for Lucky Bags},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}