关于「买手联盟的黑名单:共享覆盖率、误报与保证金」,《商帮多边惩罚与买手信任网络》给出的定量结论是什么?
结论:第五章计算买手联盟的共享黑名单:覆盖率 0.95 时可托付的订单额上限为 7.163 期租金,上报滞后一个周期就降到 3.673;误报与保证金的作用也逐一量化。
数据口径:书中代理人、订单与网络参数均为示意;全部计算由书中程序生成;本书是非个人化的经营分析,不构成经营、法律或理财建议。 第四章求临界群规模:链状熟人网络在示意参数下只有几十到一百多人,均匀混合网络要到 33742 人。
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: 0.95, 3.673, 7.163
| largest order value at stake (in units of one period's rent) | ||||
| share s | lag 0 | lag 1 | lag 3 | lag 6 |
| 0.00 | 0.818 | 0.818 | 0.818 | 0.818 |
| 0.50 | 2.077 | 1.385 | 0.933 | 0.829 |
| 0.80 | 4.263 | 2.368 | 1.132 | 0.847 |
| 0.95 | 7.163 | 3.673 | 1.396 | 0.871 |
| 1.00 | 9.000 | 4.500 | 1.564 | 0.886 |
| check, s = 0, lag 0: a r / (1 - a) with a = 0.45 is 0.818 | ||||
| full sharing, no lag: d / (1 - d) = 9.000 |
核心公式
\[E_{\max}(B) = p_f\,B + \frac{a\,(r - i\,B)}{1-a}, \qquad a = d\,(1 - H\,(1-s)).\]
查看完整验算程序
# Chapter 5. p51: what a shared blacklist buys a buyer alliance.
# Model: an agent who defaults on one buyer is reported to the shared
# list after a lag of L hiring periods; a share s of the other buyers
# consult the list. Each period the defaulter finds an opening with
# probability H = 0.5 and is taken on unless that buyer has seen the
# report: q_t = H for t <= L and H (1 - s) after. With rent r = 1 a
# period, outside value w0 = 1 and d = 0.90, we find by bisection the
# largest one-shot gain G_max (order value at stake) that still leaves
# honesty optimal. All parameters are illustrative.
# 模型:黑名单共享覆盖率 s 与上报滞后 L 决定可托付的最大订单额。
import math
R, W0, H, D = 1.0, 1.0, 0.5, 0.90
def margin(g, s, lag, horizon=400):
w = W0 + R
v = w / (1.0 - D)
q_tail = H * (1.0 - s)
u = (q_tail * v + (1.0 - q_tail) * W0) / (1.0 - (1.0 - q_tail) * D)
for t in range(horizon, 0, -1):
q = H if t <= lag else q_tail
u = q * v + (1.0 - q) * (W0 + D * u)
return v - math.fsum([w, g, D * u])
def g_max(s, lag):
lo, hi = 0.0, 1000.0
for _ in range(70):
mid = 0.5 * (lo + hi)
if margin(mid, s, lag) >= 0.0:
lo = mid
else:
hi = mid
return lo
if __name__ == "__main__":
print("largest order value at stake (in units of one period's rent)")
print(" share s lag 0 lag 1 lag 3 lag 6")
for s in (0.0, 0.5, 0.8, 0.95, 1.0):
cells = ["{:6.3f}".format(g_max(s, lag)) for lag in (0, 1, 3, 6)]
print(" {:.2f} ".format(s) + " ".join(cells))
a = D * (1.0 - H)
print("check, s = 0, lag 0: a r / (1 - a) with a = {:.2f} is {:.3f}"
.format(a, a * R / (1.0 - a)))
print("full sharing, no lag: d / (1 - d) = {:.3f}".format(D / (1 - D)))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《商帮多边惩罚声誉机制、马格里布商人代理博弈与买手社群信任网络母典》第 5 章「买手联盟的黑名单:共享覆盖率、误报与保证金」(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-222 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026222,
author = {Fan, Yafu},
title = {Multilateral Punishment in Merchant Networks: Maghribi Agency Games and Trust Networks in Buyer Communities},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}