关于「零抽成木马:何时加入是占优策略」,《无人空间计价、储值浮存金与平台收编》给出的定量结论是什么?

结论:第五章说明只要平台派单被蚕食的比例不超过 0.8105,加入就胜过自己经营,但只有平台派单低于自带流量的 1.2500 倍时才胜过统一抽成的竞争者;永久抽成会招来带客,除非发现机会超过 0.1907。

数据口径:书中的客人、需求、费率与参数均为示意模型;全部计算由书中程序生成;本书不是法律、财务、会计或税务方面的建议。 第七章读《新唐书》的飞钱、宋代出租官房的店宅务与双边市场理论:利特尔定律给出 17500,模拟为 17554。

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。 与结论吻合的打印值: 0.8105, 1.2500

a new booking keeps 0.77 of value; a moved one loses 0.18
c threshold where T stops beating staying alone: 0.8105
I/OcT - aloneT - rivalT dominates both
0.10.09240.011040.0yes
0.10.35820.010680.0yes
0.10.62400.010320.0yes
0.10.9-1020.09960.0no
0.30.027720.09120.0yes
0.30.317460.08040.0yes
0.30.67200.06960.0yes
0.30.9-3060.05880.0no
0.60.055440.06240.0yes
0.60.334920.04080.0yes
0.60.614400.01920.0yes
0.60.9-6120.0-240.0no
1.00.092400.02400.0yes
1.00.358200.0-1200.0no
1.00.624000.0-4800.0no
1.00.9-10200.0-8400.0no
c = 0.0: T beats F while I/O < 1.2500
c = 0.5: T beats F while I/O < 0.7692
c = 0.9: T beats F while I/O < 0.5882
查看完整验算程序
# Chapter 5. p09: The zero-take Trojan. A venue owner has own traffic O
# (bookings a month) at value v = TWD 2000, marginal cost 0.05 of value
# (cleaning). Offer T: free booking software, 0% on own traffic, 18% on
# orders the platform sends. The platform sends I orders; a share c of
# them are clients who would have booked directly anyway (cannibalised).
# Offer F: a rival marketplace taking 10% of every order, own traffic
# included, sending the same I. We find where T beats staying alone and
# beats F, over a grid of I/O and c.
TAKE, RIVAL, MC, V = 0.18, 0.10, 0.05, 2000.0

def stay(o):
    return o * V * (1 - MC)

def trojan(o, i, c):
    new = i * (1 - c)
    moved = i * c
    own = (o - moved) * V * (1 - MC)
    return own + (new + moved) * V * (1 - TAKE - MC)

def rival(o, i, c):
    new = i * (1 - c)
    return (o + new) * V * (1 - RIVAL - MC)

if __name__ == "__main__":
    o = 60.0
    print("a new booking keeps {:.2f} of value; a moved one loses {:.2f}"
          .format(1 - TAKE - MC, TAKE))
    print("c threshold where T stops beating staying alone: {:.4f}".format(
        (1 - TAKE - MC) / (1 - MC)))
    print("I/O    c     T - alone    T - rival   T dominates both")
    for ratio in (0.1, 0.3, 0.6, 1.0):
        for c in (0.0, 0.3, 0.6, 0.9):
            i = ratio * o
            a = trojan(o, i, c) - stay(o)
            b = trojan(o, i, c) - rival(o, i, c)
            print("{:4.1f}  {:4.1f}  {:11.1f}  {:11.1f}   {}".format(
                ratio, c, a, b, "yes" if a > 0 and b > 0 else "no"))
    # T - F = 0.10 O V - I V (0.08 + 0.10 c)
    for c in (0.0, 0.5, 0.9):
        lim = RIVAL / (TAKE - RIVAL * (1 - c))
        print("c = {:.1f}: T beats F while I/O < {:.4f}".format(c, lim))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《无人共享空间收益管理、跨店储值浮存金与木马收编博弈》第 5 章「零抽成木马:何时加入是占优策略」(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-113 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026113,
  author    = {Fan, Yafu},
  title     = {Unmanned Shared-Space Revenue Management, Cross-Venue Prepaid Float and Zero-Take-Rate Trojan Onboarding},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论