关于「超订的报童模型:边际订单法则、补偿成本与成团爽约的相关性陷阱」,《摄影场馆的装箱、退改与折旧》给出的定量结论是什么?

结论:第四章把超订写成报童问题,多接一单多赚 679 元,但爽约的日内相关达到 0.15 时就不应超订。

数据口径:书中参数为示意;全部计算由书中程序生成;本书不是个人化的税务、法律或投资建议。 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。

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: 679

extra kP(any bump)E[bumped]E[profit]changeP(room)
00.00000.000017587+00.6404
10.31650.316518267+6790.2951
20.65830.936817487-7790.1087
30.86341.721115921-15670.0341
40.95362.571114039-18820.0095
best k = 1: 18267 yuan vs 17587 with no overbooking
rule: add a booking while P(room) > 0.4795
P(room) = chance the extra booking finds an empty slot
查看完整验算程序
# Chapter 4. p41: overbooking a peak Saturday as a newsvendor problem.
# 8 half-day slots at 2400 yuan.  The venue takes 8 + k bookings; each
# client shows independently with probability 0.88.  A no-show forfeits
# a deposit of 0.3 of the price.  A client who shows but finds no slot
# gets her money back plus compensation of 1.0 price (a free future
# session at cost plus goodwill).  Exact binomial expectation.
# Newsvendor reading: book one more while the chance that the extra
# booking lands on an empty slot beats the cost ratio.
# Illustrative parameters.
import math

C, P, SHOW, DEP, COMP = 8, 2400, 0.88, 0.3, 1.0

def binom(n, k, p):
    return math.comb(n, k) * p ** k * (1 - p) ** (n - k)

def profit(n):
    parts, pb, eb = [], [], []
    for s in range(n + 1):
        w = binom(n, s, SHOW)
        bump = max(0, s - C)
        parts.append(w * (min(s, C) * P + (n - s) * DEP * P
                          - bump * COMP * P))
        pb.append(w if bump else 0.0)
        eb.append(w * bump)
    return math.fsum(parts), math.fsum(pb), math.fsum(eb)

def room(n):
    return math.fsum(binom(n, s, SHOW) for s in range(min(n, C - 1) + 1))

need = (COMP * SHOW - (1 - SHOW) * DEP) / (SHOW * (1 + COMP))
print("extra k  P(any bump)  E[bumped]  E[profit]  change  P(room)")
prev = None
best = None
for k in range(0, 5):
    v, pb, eb = profit(C + k)
    ch = 0 if prev is None else v - prev
    print("{:7d}  {:11.4f}  {:9.4f}  {:9.0f}  {:+6.0f}  {:7.4f}".format(
        k, pb, eb, v, ch, room(C + k)))
    prev = v
    if best is None or v > best[1]:
        best = (k, v)
print("best k = {}: {:.0f} yuan vs {:.0f} with no overbooking".format(
    best[0], best[1], profit(C)[0]))
print("rule: add a booking while P(room) > {:.4f}".format(need))
print("P(room) = chance the extra booking finds an empty slot")

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《多规格实景空间时段一维装箱、动态退改期权阶梯与高损耗器材加速折旧运筹》第 4 章「超订的报童模型:边际订单法则、补偿成本与成团爽约的相关性陷阱」(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-173 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026173,
  author    = {Fan, Yafu},
  title     = {One-Dimensional Time-Slot Packing for Multi-Format Studio Sets, Dynamic Cancellation-Option Ladders and Accelerated Depreciation of High-Wear Equipment},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论