关于「原典考辨:天行有常、守株待兔、有利之物与服从自然」,《效用、真实与双账本》给出的定量结论是什么?

结论:只看 3 次决策,C 在 0.528 的路径上领先 B,平均财富是 B 的 1.185 倍,中位财富是 1.02 倍:短期看,单账本似乎更有利。看 10 次,C 只在 0.409 的路径上领先,平均财富却是 1.689 倍,中位财富已降到 0.7993。看 30 次,领先比例 0.339,平均财富 5.931 倍,中位 0.5479 倍。看 100 次,领先比例 0.229,平均财富 44.62 倍,中位 0.1737 倍。看 300 次,C 只在 0.097 的路径上领先,平均财富却是 B 的 864 倍,中位财富只剩 0.002656 倍。

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: 0.002656, 0.097, 0.1737, 0.229, 0.339, 0.409

HC ahead of Bmean ratio C/Bmedian ratio C/B
30.5281.1851.02
100.4091.6890.7993
300.3395.9310.5479
1000.22944.620.1737
3000.0978640.002656
查看完整验算程序
# Chapter 7. p73: "expedient" over a short run and over the long run
# Reading of William James, "Pragmatism", Lecture VI: truth judged by
# what works in our thinking, and by what works in the long run. The
# p62 ventures (2,000 paths, success returns 1.5 times the stake, frame
# lift 0.02 on both policies). B keeps the probability ledger calibrated
# (Kelly of q); C writes the frame into the ledger (Kelly of q + 0.12).
# For each horizon H: the share of paths on which C is ahead of B, the
# ratio of MEAN wealth (C over B) and the ratio of MEDIAN wealth. A
# belief that "works" in the mean over a few rounds can lose on almost
# every path over many. Illustrative only.
import math
import random
import statistics

ODDS, LIFT, INFL, N, PATHS = 1.5, 0.02, 0.12, 300, 2000
rng = random.Random(42773)
QS = [[rng.uniform(0.40, 0.60) for _ in range(N)] for _ in range(PATHS)]
US = [[rng.random() for _ in range(N)] for _ in range(PATHS)]


def kelly(p):
    return min(0.99, max(0.0, (p * (ODDS + 1) - 1) / ODDS))


def logs(infl, qs, us):
    out = []
    for q, u in zip(qs, us):
        f = kelly(min(0.99, q + infl))
        out.append(math.log(1 + ODDS * f) if u < q + LIFT
                   else math.log(1 - f))
    return out


B = [logs(0.0, q, u) for q, u in zip(QS, US)]
C = [logs(INFL, q, u) for q, u in zip(QS, US)]
print("   H   C ahead of B   mean ratio C/B   median ratio C/B")
for h in (3, 10, 30, 100, 300):
    lb = [math.fsum(x[:h]) for x in B]
    lc = [math.fsum(x[:h]) for x in C]
    ahead = len([1 for a, b in zip(lc, lb) if a > b]) / PATHS
    mean_r = (statistics.fmean([math.exp(x) for x in lc])
              / statistics.fmean([math.exp(x) for x in lb]))
    med_r = math.exp(statistics.median(lc) - statistics.median(lb))
    print("%4d  %13.3f  %15.4g  %17.4g" % (h, ahead, mean_r, med_r))

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 7 章「原典考辨:天行有常、守株待兔、有利之物与服从自然」(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