What does Partner Selection on the Pareto Frontier conclude, in numbers, about “Berkson Selection: Where the Observed Trade-off Comes From”?

Answer: Chapter 2 is the core of the book: starting from two independent traits and keeping only the top 10 % by overall appeal, the correlation becomes −0.708, and market clearing produces the same trade-off.

Scope of the numbers: Its populations, traits and parameters are synthetic or illustrative, and all calculations are generated by its programs. Marked boxes separate evidence, hypotheses and classical sources; the book contains 22 programs.

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: 10

people per side: 2,000
all partners: corr(T, E) = +0.004
chooser decilemean Tmean Ecorr(T, E) among partners
1-1.17-1.18-0.540
2-0.70-0.71-0.742
3-0.49-0.37-0.751
4-0.24-0.23-0.782
5-0.14+0.00-0.864
6-0.03+0.23-0.759
7+0.30+0.27-0.828
8+0.48+0.43-0.817
9+0.68+0.76-0.794
10+1.21+1.21-0.598
within any band the two traits trade off; the market, not the
traits, produces the negative slope
Show the full program
# Chapter 2. p22: market clearing makes the obtainable set trade off
# Two sides of 2000 people each. Everyone has independent traits
# T (tension signal) and E (exclusive investment), and overall
# desirability D = T + E + noise. With common preferences the stable
# matching is assortative on D (Becker 1973; Kalick and Hamilton 1986).
# A chooser in a given band of own D obtains partners whose D sits in a
# narrow band, so among those partners T and E trade off one for one.
import math
import random

rng = random.Random(41302)
N = 2000


def person():
    t, e = rng.gauss(0.0, 1.0), rng.gauss(0.0, 1.0)
    return (t + e + rng.gauss(0.0, 0.5), t, e)


def corr(xs, ys):
    n = len(xs)
    mx, my = math.fsum(xs) / n, math.fsum(ys) / n
    sxy = math.fsum((x - mx) * (y - my) for x, y in zip(xs, ys))
    sxx = math.fsum((x - mx) ** 2 for x in xs)
    syy = math.fsum((y - my) ** 2 for y in ys)
    return sxy / math.sqrt(sxx * syy)


side_a = sorted(person() for _ in range(N))
side_b = sorted(person() for _ in range(N))
pairs = list(zip(side_a, side_b))  # rank i of A with rank i of B
ta = [p[1][1] for p in pairs]
ea = [p[1][2] for p in pairs]
print("people per side: {:,}".format(N))
print("all partners: corr(T, E) = {:+.3f}".format(corr(ta, ea)))
print("chooser decile  mean T  mean E  corr(T, E) among partners")
for dec in range(10):
    blk = pairs[dec * N // 10:(dec + 1) * N // 10]
    ts = [p[1][1] for p in blk]
    es = [p[1][2] for p in blk]
    print("{:>14}  {:+.2f}   {:+.2f}   {:+.3f}".format(
        dec + 1, round(math.fsum(ts) / len(ts), 2) + 0.0,
        round(math.fsum(es) / len(es), 2) + 0.0,
        corr(ts, es)))
print("within any band the two traits trade off; the market, not the")
print("traits, produces the negative slope")

Single source and citation

Fan, Y. (2026). Dual-Mating Multi-Objective Conflicts, Parental-Investment Asymmetry and the Pareto Frontier of Partner Selection (Version 1.0.0), Chapter 2: Berkson Selection: Where the Observed Trade-off Comes From. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-421 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026421,
  author    = {Fan, Yafu},
  title     = {Dual-Mating Multi-Objective Conflicts, Parental-Investment Asymmetry and the Pareto Frontier of Partner Selection},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book