关于「自我实现增益的上限:信心能抬高多少成功率」,《效用、真实与双账本》给出的定量结论是什么?

结论:GAMMA 为 0.02 时,最优膨胀只有 0.10,最大增益 0.0094,远低于这一 GAMMA 下的天花板 0.20。b 一旦到 0.5,结果已经是 -0.0776;到 1,是 -0.2863。GAMMA 为 0.05 时,最优膨胀是 0.25,最大增益 0.0537;b 为 0.5 时还剩 0.0224,b 为 1 时已经是 -0.1363。

数据口径:书中参数与数据均为示意或合成,全部计算由书中程序生成;本书不是医疗建议,不提供任何临床心理服务,不是个人投资建议,不承诺任何收益;功能性信念只用于自己的行动框架,不用于欺骗他人。 第六章建立双账本:行动框架可按效用编码,概率账以 Brier 分数校准,硬边界账不让步;框架改写概率账时,示意破产比例从 0.0925 升到 0.7055。

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: -0.0776, -0.1363, -0.2863, 0.0094, 0.02, 0.0224

calibrated effort 1.833, true value 3.1674
gammabest bmax gainbound at bceilinggain b=0.5gain b=1
0.000.000.00000.00000.00-0.1443-0.3863
0.020.100.00940.01820.20-0.0776-0.2863
0.050.250.05370.10000.500.0224-0.1363
0.100.500.18910.33331.000.18910.1137
0.150.750.38080.64291.500.35570.3637
查看完整验算程序
# Chapter 4. p41: the self-fulfilling gain has a ceiling and a peak
# One task. Effort e costs KAPPA per unit; success is worth V. The true
# success probability is p(e) = P0 + (PMAX - P0)(1 - exp(-e / TAU)),
# plus a direct confidence channel GAMMA * b / (1 + b) (persistence,
# composure) that saturates at GAMMA, capped at 1. An agent with
# confidence inflation b believes its effort works (1 + b) times better
# than it does and chooses effort to maximise its BELIEVED value. Gain
# G(b) = true value at b minus true value of a calibrated agent (b = 0).
# Bound: G(b) <= V * GAMMA * b / (1 + b) < V * GAMMA, because the effort
# channel can only lose against the calibrated optimum. Illustrative.
import math

V, KAPPA, P0, PMAX, TAU = 10.0, 1.0, 0.2, 0.7, 2.0


def effort(b):
    k = V * (PMAX - P0) * (1 + b) / TAU
    return max(0.0, TAU * math.log(k / KAPPA)) if k > KAPPA else 0.0


def true_value(b, gamma):
    e = effort(b)
    p = P0 + (PMAX - P0) * (1 - math.exp(-e / TAU)) + gamma * b / (1 + b)
    return V * min(1.0, p) - KAPPA * e


print("calibrated effort %.3f, true value %.4f" % (
    effort(0), true_value(0, 0)))
print("gamma  best b  max gain  bound at b  ceiling  gain b=0.5  gain b=1")
for gamma in (0.0, 0.02, 0.05, 0.1, 0.15):
    grid = [i / 100 for i in range(0, 301)]
    base = true_value(0, gamma)
    best = max(grid, key=lambda b: (true_value(b, gamma), -b))
    g_best = true_value(best, gamma) - base
    print("%5.2f  %6.2f  %8.4f  %10.4f  %7.2f  %10.4f  %8.4f" % (
        gamma, best, g_best, V * gamma * best / (1 + best), V * gamma,
        true_value(0.5, gamma) - base, true_value(1.0, gamma) - base))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《效用优先于字面真实:演化适存度博弈、功能性信念重塑与物理硬边界母典》第 4 章「自我实现增益的上限:信心能抬高多少成功率」(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-427 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026427,
  author    = {Fan, Yafu},
  title     = {When Utility Outperforms Literal Truth: Evolutionary Fitness-Payoff Games, Functional Belief Engineering and Physical Guardrails},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论