What does Password Entropy and Least Privilege conclude, in numbers, about “The Forced-Rotation Paradox: Predictable Variants and an Overrated Attack Window”?
Answer: Chapter 1 shows that an attacker who needs one day of access still succeeds with probability 0.991 under a 90-day forced change, corrects the blueprint's "no longer recommends" to the "shall not" of Revision 4, and sets out the counter-evidence of a study of self-reported behaviour.
Scope of the numbers: Its firms, accounts and probabilities are illustrative, and all calculations are generated by its programs; it is not personalised business, legal, financial or compliance advice, and its audit routine and tool never touch a real password. Chapter 6 uses minimum role covers and the recovery graph to show that the illustrative firm expects to lose 0.4972 accounts a year, and 0.1899 after hardening the main mailbox alone.
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: 0.991, 90
| P(attacker keeps access for D days), compromise at random time | ||||
| online q=0.17 | ||||
| rotate every | D=1 | D=7 | D=30 | D=90 |
| never | 1.000 | 1.000 | 1.000 | 1.000 |
| 180 days | 0.995 | 0.968 | 0.862 | 0.585 |
| 90 days | 0.991 | 0.935 | 0.723 | 0.170 |
| 30 days | 0.972 | 0.806 | 0.170 | 0.005 |
| offline q=0.41 | ||||
| rotate every | D=1 | D=7 | D=30 | D=90 |
| never | 1.000 | 1.000 | 1.000 | 1.000 |
| 180 days | 0.997 | 0.977 | 0.902 | 0.705 |
| 90 days | 0.993 | 0.954 | 0.803 | 0.410 |
| 30 days | 0.980 | 0.862 | 0.410 | 0.069 |
| with persistence s=0.5 (forwarding rule survives a change) | ||||
| 90 days | 0.995 | 0.968 | 0.862 | 0.585 |
| 30 days | 0.986 | 0.903 | 0.585 | 0.502 |
Key formula
Show the full program
# Chapter 1. p11: does forced rotation shorten a stolen password's life?
# Model: the attacker obtains a password at a random moment; the time U
# to the next forced change is uniform on [0, T). After each change the
# user picks, with probability q, a variant derivable from the old one,
# so the attacker gets back in. The attacker needs D days of access.
# q = 0.17 is the online result (under 5 guesses) and q = 0.41 the
# offline result of Zhang, Monrose and Reiter (2010), from a dataset of
# over 7700 accounts (sourced); the persistence probability s is
# illustrative: a forwarding rule or app token planted after entry that
# survives the password change.
# Exact: P = s + (1 - s) * sum_k (1 - q) q^k P(U >= D - kT).
import math
Q_ONLINE = 0.17 # Zhang et al. 2010, under 5 online guesses
Q_OFFLINE = 0.41 # Zhang et al. 2010, offline, under 3 s per account
def p_reach(days, period, q, persist=0.0, kmax=400):
"""Probability that the attacker holds access for `days` days."""
if period is None:
return 1.0
terms = []
for k in range(kmax):
need = days - k * period # U must be at least this
frac = min(1.0, max(0.0, 1.0 - need / period))
terms.append((1.0 - q) * q ** k * frac)
return persist + (1.0 - persist) * math.fsum(terms)
DAYS = (1, 7, 30, 90)
print("P(attacker keeps access for D days), compromise at random time")
for q, tag in ((Q_ONLINE, "online q=0.17"), (Q_OFFLINE, "offline q=0.41")):
print(tag)
print(" rotate every " + "".join("D={:<5d}".format(d) for d in DAYS))
for per in (None, 180, 90, 30):
lab = "never" if per is None else "{} days".format(per)
row = "".join("{:<7.3f}".format(p_reach(d, per, q)) for d in DAYS)
print(" {:<14s}{}".format(lab, row))
print("with persistence s=0.5 (forwarding rule survives a change)")
for per in (90, 30):
row = "".join("{:<7.3f}".format(p_reach(d, per, Q_ONLINE, 0.5))
for d in DAYS)
print(" {:<14s}{}".format("{} days".format(per), row))Single source and citation
Fan, Y. (2026). Password Entropy and Credential Stuffing, the Forced-Rotation Paradox and Least-Privilege Accounts for One-Person Firms (Version 1.0.0), Chapter 1: The Forced-Rotation Paradox: Predictable Variants and an Overrated Attack Window. 5D Library. DOI pending
DOI pending · monograph page goes live after publication · 5DL-B-218 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026218,
author = {Fan, Yafu},
title = {Password Entropy and Credential Stuffing, the Forced-Rotation Paradox and Least-Privilege Accounts for One-Person Firms},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}Other chapter answers in this book
- Chapter 2 · The Entropy Ledger: log2(N^L), Random Long Passphrases and Complex Short Passwords
- Chapter 3 · Guessing under a Zipf Law: Why Shannon Entropy Overrates Human-Chosen Passwords
- Chapter 4 · The Credential-Stuffing Funnel: Reuse Rates, Breach Corpora and Second Factors
- Chapter 5 · Throttling, Blocklists and Hash Prefixes: the Ceiling on Online Guessing
- Chapter 6 · The Permission Graph and its Minimum Cover: a One-Person Firm's Blast Radius
- Chapter 8 · The Permission-Matrix Audit Sheet: an Empirical Loop and a Falsification Plan