What does First-Move Probes and Hidden Market Truth conclude, in numbers, about “Bringing Willingness to Pay into View, and a Counterpart's Floor: BDM, Hypothetical Bias and Inference from Public Information”?
Answer: Chapter 6 covers the BDM mechanism, hypothetical bias between stated and real purchases, and inference of a counterpart's floor from public rate cards and public open calls only, with a posterior interval from 400 to 480.
Scope of the numbers: Parameters are illustrative, calculations come from the book's programs, and all data are synthetic. This is general business analysis, not personalised legal, tax or investment advice; it promises no income.
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: 400, 480
| public open calls (budget, invited, yes): | ||
| (300, 40, 6), (500, 40, 26), (800, 40, 39) | ||
| rate cards: n = 12, median posted 816 | ||
| source | mean M | 90% interval |
| cards only | 496.2 | [400, 600] |
| calls only | 429.3 | [390, 470] |
| both | 438.8 | [400, 480] |
Show the full program
# Chapter 6. p63: reading a counterpart tier's reserve from public data.
# Illustrative, synthetic. Creators in one tier have reserve fees
# r ~ log-normal(median M, log spread 0.4); M is unknown (grid 200 to
# 900 units, flat prior). Two public sources only, no posing as a client:
# (1) public rate cards: a posted fee = r * markup, markup uniform on
# 1.1 to 1.5 (unknown per creator);
# (2) public open calls: we post a real, payable budget b to 40 eligible
# creators; a creator replies yes when b >= r.
# The data are drawn once from M = 450; the posterior of M is reported
# for each source alone and for both together.
import math
import random
GRID = list(range(200, 901, 10))
SIG = 0.4
MARKUPS = [1.1, 1.2, 1.3, 1.4, 1.5]
def cdf(x, m):
return 0.5 * math.erfc(-(math.log(x) - math.log(m)) / (SIG * 2 ** 0.5))
def pdf(x, m):
z = (math.log(x) - math.log(m)) / SIG
return math.exp(-0.5 * z * z) / (x * SIG * math.sqrt(2 * math.pi))
def loglik_cards(cards, m):
out = []
for c in cards:
dens = math.fsum(pdf(c / k, m) / k for k in MARKUPS) / len(MARKUPS)
out.append(math.log(dens))
return math.fsum(out)
def loglik_calls(calls, m):
out = []
for b, n, yes in calls:
p = min(max(cdf(b, m), 1e-12), 1 - 1e-12)
out.append(yes * math.log(p) + (n - yes) * math.log(1 - p))
return math.fsum(out)
def summarise(logs):
top = max(logs)
w = [math.exp(x - top) for x in logs]
tot = math.fsum(w)
w = [x / tot for x in w]
mean = math.fsum(g * x for g, x in zip(GRID, w))
lo, hi = None, None
for i, g in enumerate(GRID):
cum = math.fsum(w[:i + 1])
if lo is None and cum >= 0.05:
lo = g
if hi is None and cum >= 0.95:
hi = g
return mean, lo, hi
if __name__ == "__main__":
rng = random.Random(39563)
true_m = 450.0
cards = [math.exp(rng.gauss(math.log(true_m), SIG))
* rng.choice(MARKUPS) for _ in range(12)]
calls = []
for b in (300, 500, 800):
yes = 0
for _ in range(40):
if math.exp(rng.gauss(math.log(true_m), SIG)) <= b:
yes += 1
calls.append((b, 40, yes))
print("public open calls (budget, invited, yes):")
print(" " + ", ".join("({}, {}, {})".format(*c) for c in calls))
print("rate cards: n = {}, median posted {:.0f}".format(
len(cards), sorted(cards)[6]))
print(" source mean M 90% interval")
for name, f in (("cards only", lambda m: loglik_cards(cards, m)),
("calls only", lambda m: loglik_calls(calls, m)),
("both", lambda m: loglik_cards(cards, m)
+ loglik_calls(calls, m))):
mean, lo, hi = summarise([f(m) for m in GRID])
print(" {:<12} {:>7.1f} [{}, {}]".format(name, mean, lo, hi))Single source and citation
Fan, Y. (2026). High-Agency Active Inference, Intentional Perturbation Probes and Hidden Market-Truth Exploration (Version 1.0.0), Chapter 6: Bringing Willingness to Pay into View, and a Counterpart's Floor: BDM, Hypothetical Bias and Inference from Public Information. 5D Library. DOI pending
DOI pending · monograph page goes live after publication · 5DL-B-395 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026395,
author = {Fan, Yafu},
title = {High-Agency Active Inference, Intentional Perturbation Probes and Hidden Market-Truth Exploration},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}Other chapter answers in this book
- Chapter 1 · The Bias Floor of Confounding: Why a Longer Passive Log Is Wrong with More Confidence
- Chapter 2 · Identifiability: What the Back Door, the Instrument and the Published Threshold Each Require
- Chapter 3 · Expected Information Gain: Put the Probe Where It Teaches the Most
- Chapter 4 · Exploration and Exploitation: The Regret Ledger of Thompson Sampling and Fixed Habits
- Chapter 5 · Commitment Strength Made Operational: Probes per Week, Delay to the First Probe and the Stopping Rule
- Chapter 7 · The Classics Examined: Lu You, Wang Yangming, Bacon and Bernard on Doing and on Provoked Observation
- Chapter 8 · Field Tools and Checker Specifications: The First-Move Probe Ledger, Pre-Registered Tests and the Plan for Refutation