关于「超售:航空公司的算术搬进美容室」,《工作室超售、储值浮存金与升单阶梯》给出的定量结论是什么?

结论:第二章把超售写成报童问题:让到场客户扑空的代价为价格 3 倍时最优只多排 1 位,每天多 0.033 个满日收入;只在两位以上高风险客户的日子超售,净收入从 6.717 升到 6.801。

数据口径:书中的客户、价格、比率与成本均为示意模型;全部计算由书中程序生成;本书不是财务、会计、法律或医疗建议,不对任何疗程或服务的效果作出主张。 第八章把这些合成一周的经营与一份检验方案,并给出一份清单。

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: 6.717, 6.801

policychairs usedbumps per 100 daysnet per day
none0.8400.06.717
blind0.91820.46.734
per-flag0.95656.25.966
two-flags0.90615.06.801
two-flags over none: 0.084 a day, 0.0125 of the day's net
查看完整验算程序
# Chapter 2. p04: Overbook everyone, or only the risky? Clients come in
# two kinds: 70 percent miss with probability 0.06 and 30 percent with
# 0.40 (mean 0.162). The studio flags a client as risky when she missed
# at least one of her last three bookings. Each day 8 slots are booked.
# Policies: none; one blind extra booking every day; one extra booking
# for each flagged client (at most two); one extra booking only on days
# with two or more flagged clients. Bump cost 3, price 1. 20,000
# simulated days per policy, seed 107.
import random

rng = random.Random(107)
DAYS, S, B = 20000, 8, 3.0

def client():
    p = 0.40 if rng.random() < 0.30 else 0.06
    flag = any(rng.random() < p for _ in range(3))
    return p, flag

def day(policy):
    cl = [client() for _ in range(S)]
    extra = 0
    if policy == "blind":
        extra = 1
    elif policy == "per-flag":
        extra = min(2, sum(f for _, f in cl))
    elif policy == "two-flags":
        extra = 1 if sum(f for _, f in cl) >= 2 else 0
    cl += [client() for _ in range(extra)]
    shows = sum(rng.random() >= p for p, _ in cl)
    return min(shows, S), max(shows - S, 0)

if __name__ == "__main__":
    print("   policy      chairs used   bumps per 100 days   net per day")
    nets = {}
    for pol in ("none", "blind", "per-flag", "two-flags"):
        used = bumps = 0
        for _ in range(DAYS):
            u, b = day(pol)
            used += u
            bumps += b
        net = (used - B * bumps) / DAYS
        nets[pol] = net
        print("   {:9s}   {:11.3f}   {:18.1f}   {:11.3f}".format(
            pol, used / DAYS / S, 100 * bumps / DAYS, net))
    print("   two-flags over none: {:.3f} a day, {:.4f} of the day's net"
          .format(nets["two-flags"] - nets["none"],
                  (nets["two-flags"] - nets["none"]) / nets["none"]))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《预约制工作室超售排队、储值沉淀资金池与高客单升单阶梯》第 2 章「超售:航空公司的算术搬进美容室」(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-107 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026107,
  author    = {Fan, Yafu},
  title     = {Appointment Studio Overbooking Queues, Prepaid Float Dynamics and High-Ticket Ascension Ladders},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论