What does Critical Fractiles for Pre-Season Buying conclude, in numbers, about “The Critical Fractile: Cost Ratios, Not Averages, Set the Buy”?

Answer: Chapter 1 derives the critical fractile from marginal arithmetic; under illustrative parameters the optimal order equals the mean only at a margin of exactly 0.3750, and ordering the mean for an eight-tenths-margin item loses 6.96 % of expected profit. Reconstructed from published data of the Schweitzer and Cachon experiment, participants pulled their orders 0.644 (high profit) and 0.787 (low profit) of the way toward the mean.

Scope of the numbers: Its goods, stores and demand parameters are illustrative; all calculations are generated by its programs; the book is not individual business advice. Chapter 1 derives the critical fractile from marginal arithmetic; under illustrative parameters the optimal order equals the mean only at a margin of exactly 0.3750, and ordering the mean for an eight-tenths-margin item loses 6.96 % of expected profit.

中文版

Data table

The numbers below are printed verbatim by the chapter's Python program and can be reproduced by running it. standard-library Python onlyprinted values that match the answer: 0.644, 0.787

condCFq*meanobservedpull aG(obs)/G*G(mean)/G*
high0.75225.0150.0176.680.6441.13841.3333
low0.2575.0150.0134.060.7871.20671.3333
under-order high: 48.32 units; over-order low: 59.06 units
high order is 26.68 above the mean; low order 15.94 below it

Key formula

\[q^* = F^{-1}\!\left(\frac{c_u}{c_u + c_o}\right).\]
Show the full program
# Chapter 1. p12: Pull-to-center in the Schweitzer-Cachon (2000) data.
# Reported: optimal orders 225 (high profit, CF 0.75) and 75 (low
# profit, CF 0.25); average orders 176.68 and 134.06 (as quoted in the
# Management Science replication report). Demand is uniform; the two
# optima imply an interval [0, 300] with mean 150 (our inference).
# Anchoring model: q = a * mean + (1 - a) * q*, solved for a.
# Mismatch cost G(q) = cu E(D-q)+ + co E(q-D)+ under U(0, L), exact.
import math

L = 300.0
MEAN = L / 2.0
CASES = (("high", 0.75, 225.0, 176.68), ("low", 0.25, 75.0, 134.06))


def mismatch(q, cu, co):
    under = (L - q) ** 2 / (2.0 * L)
    over = q ** 2 / (2.0 * L)
    return math.fsum([cu * under, co * over])


print("cond  CF    q*     mean   observed  pull a  G(obs)/G*  G(mean)/G*")
for name, cf, qs, obs in CASES:
    cu, co = cf, 1.0 - cf
    a = (qs - obs) / (qs - MEAN)
    g0 = mismatch(qs, cu, co)
    fmt = "{:5} {:.2f}  {:5.1f}  {:5.1f}  {:8.2f}  {:.3f}   {:.4f}"
    r1 = mismatch(obs, cu, co) / g0
    r2 = mismatch(MEAN, cu, co) / g0
    print((fmt + "     {:.4f}").format(name, cf, qs, MEAN, obs, a, r1, r2))
gap_hi = 225.0 - 176.68
gap_lo = 134.06 - 75.0
print("under-order high: {:.2f} units; over-order low: {:.2f} units"
      .format(gap_hi, gap_lo))
print("high order is {:.2f} above the mean; low order {:.2f} below it"
      .format(176.68 - MEAN, MEAN - 134.06))

Single source and citation

Fan, Y. (2026). Newsvendor Critical Fractiles for Pre-Season Buying, Dirichlet Size-Curve Priors and Dynamic-Programming End-of-Season Markdowns (Version 1.0.0), Chapter 1: The Critical Fractile: Cost Ratios, Not Averages, Set the Buy. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-211 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026211,
  author    = {Fan, Yafu},
  title     = {Newsvendor Critical Fractiles for Pre-Season Buying, Dirichlet Size-Curve Priors and Dynamic-Programming End-of-Season Markdowns},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book