关于「羽翼丰满之后:外部选项、纳什分成与阶梯合约」,《达人经纪的幂律、抽查与买断》给出的定量结论是什么?
结论:第四章用纳什讨价还价解与外部选项原理说明,固定 0.60 的分成在粉丝数约 204192 时破裂,阶梯分成把模拟名册的佣金现值从 830.56 百万元提高到 1253.26 百万元。
数据口径:书中参数为示意;全部计算由书中程序生成;本书不是个人化的法律或理财建议。 第六章把阶梯分成、买断金与抽查率合成三联定价试算器,买断金取剩余佣金的 1.712 倍。
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: 0.60, 1253.26, 830.56
| edge of tier | ladder share | outside share at edge |
|---|---|---|
| 200000 | 0.60 | 0.597 |
| 500000 | 0.74 | 0.708 |
| 1000000 | 0.81 | 0.783 |
| 3000000 | 0.90 | 0.896 |
| contract walked away agency PV (m yuan) from top 1% per head | ||
| flat 0.60 449 830.56 0.014 0.601 | ||
| ladder 0 1253.26 0.055 3.472 | ||
查看完整验算程序
# Chapter 4. p42: a flat split against a ladder designed in advance.
# Same revenue and outside-option curves as p41. 2000 creators start
# near 20000 followers (lognormal spread 0.6); log followers move each
# month by a normal step (drift 0.01, volatility 0.30) for a 36-month
# contract. A creator walks away in the first month in which her
# contract share times revenue falls below her outside option. The
# agency's income is (1 - share) x revenue, discounted 0.99 a month,
# while the contract lasts. The ladder equals the flat 0.60 below
# 200000 followers and then steps up so that each tier stays just above
# the outside option at the tier's upper edge. Illustrative inputs.
import math
import random
F_LO, F_HI, T_LO, T_HI, OWN_COST = 10000, 3000000, 0.35, 0.90, 8000
LADDER = ((200000, 0.60), (500000, 0.74),
(1000000, 0.81), (3000000, 0.90), (10 ** 12, 0.92))
def revenue(f):
return 6 * f ** 0.85
def outside(f):
u = math.log(f / F_LO) / math.log(F_HI / F_LO)
th = T_LO + (T_HI - T_LO) * min(1.0, max(0.0, u))
return max(0.0, th * revenue(f) - OWN_COST)
def ladder(f):
for edge, s in LADDER:
if f < edge:
return s
return LADDER[-1][1]
def simulate(rule):
rng = random.Random(20261112)
res = []
for _ in range(2000):
x = math.log(20000) + rng.gauss(0.0, 0.6)
pv, left, peak = [], False, 0.0
for t in range(36):
x += rng.gauss(0.01, 0.30)
f = math.exp(x)
peak = max(peak, f)
s = rule(f)
if s * revenue(f) < outside(f):
left = True
break
pv.append((1 - s) * revenue(f) * 0.99 ** t)
res.append((peak, math.fsum(pv), left))
return res
print("edge of tier ladder share outside share at edge")
for edge, s in LADDER[:-1]:
print("{:12d} {:13.2f} {:22.3f}".format(edge, s,
outside(edge) / revenue(edge)))
print("contract walked away agency PV (m yuan) from top 1% per head")
for name, rule in (("flat 0.60", lambda f: 0.60), ("ladder", ladder)):
res = simulate(rule)
res.sort(key=lambda r: -r[0])
tot = math.fsum(r[1] for r in res)
top = math.fsum(r[1] for r in res[:20])
walked = len([r for r in res if r[2]])
print("{:10s} {:12d} {:19.2f} {:12.3f} {:9.3f}".format(
name, walked, tot / 1e6, top / tot, top / 20 / 1e6))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《网红经纪机构孵化幸存者幂律、达人飞单抽样审计博弈与违约买断期权定价》第 4 章「羽翼丰满之后:外部选项、纳什分成与阶梯合约」(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-172 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026172,
author = {Fan, Yafu},
title = {Survivor Power Laws in Influencer-Agency Incubation, Sampling-Audit Games against Off-Platform Deals and Option Pricing of Contract Buyouts},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}