What does Habit Curves and the 21-Day Myth conclude, in numbers, about “The Asymptotic Curve: Plateau, Gap, Rate and T95 in Closed Form”?
Answer: Chapter 2 derives T95 in closed form and shows that "21 days" assumes the typical person learns 3.14 times faster.
Scope of the numbers: Its population parameters, forgetting rates and cue indices are illustrative; all calculations are generated by its programs; it is not medical or psychological advice, nor personalised business advice. Marked boxes separate published evidence, runnable Python calculations (19 programs), classical sources and the author's hypotheses.
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: 21, 3.14, 95
| b / a | 0.857 ln(20 b / a) = 2.8416 |
| rate for T95 | 66 days: c = 0.04305 per day |
| rate for T95 | 21 days: c = 0.13531 per day (x 3.14) |
| half-gap time at c66 | 16.1 days |
| day A(t), T95 | 66 share of a A(t), T95=21 share of a |
| 0 | 5.00 0.143 5.00 0.143 |
| 7 | 12.81 0.366 23.37 0.668 |
| 21 | 22.85 0.653 33.25 0.950 |
| 42 | 30.08 0.859 34.90 0.997 |
| 66 | 33.25 0.950 35.00 1.000 |
| 84 | 34.19 0.977 35.00 1.000 |
| 120 | 34.83 0.995 35.00 1.000 |
| T95 if the start is 0 (b = a) | 69.6 days at c66 |
Key formula
Show the full program
# Chapter 2. p21: The asymptotic curve A(t) = a - b exp(-c t) in closed form.
# a is the plateau of automaticity, b the gap between plateau and start,
# c the rate per day. A(t) reaches 95 % of a when b exp(-c t) = 0.05 a,
# so T95 = ln(20 b / a) / c, and the half-gap time is ln 2 / c.
# Scale: 7 SRHI items scored 0 to 6, total 0 to 42, as in Lally et al.
# (2010); plateau a = 35 and start 5 are illustrative, near their median.
import math
A_PLAT, START = 35.0, 5.0
B = A_PLAT - START
def t95(a, b, c):
return math.log(20.0 * b / a) / c
def rate_for(days, a, b):
return math.log(20.0 * b / a) / days
c66 = rate_for(66.0, A_PLAT, B)
c21 = rate_for(21.0, A_PLAT, B)
print("b / a = %.3f ln(20 b / a) = %.4f" % (B / A_PLAT,
math.log(20 * B / A_PLAT)))
print("rate for T95 = 66 days: c = %.5f per day" % c66)
print("rate for T95 = 21 days: c = %.5f per day (x %.2f)"
% (c21, c21 / c66))
print("half-gap time at c66: %.1f days" % (math.log(2) / c66))
print("")
print(" day A(t), T95=66 share of a A(t), T95=21 share of a")
for d in (0, 7, 21, 42, 66, 84, 120):
x = A_PLAT - B * math.exp(-c66 * d)
y = A_PLAT - B * math.exp(-c21 * d)
print(" %3d %12.2f %10.3f %12.2f %10.3f"
% (d, x, x / A_PLAT, y, y / A_PLAT))
print("")
print("T95 if the start is 0 (b = a): %.1f days at c66"
% t95(A_PLAT, A_PLAT, c66))Single source and citation
Fan, Y. (2026). Asymptotic Habit-Formation Curves, Debunking the 21-Day Myth and a Cue-Based Automaticity Index (Version 1.0.0), Chapter 2: The Asymptotic Curve: Plateau, Gap, Rate and T95 in Closed Form. 5D Library. DOI pending
DOI pending · monograph page goes live after publication · 5DL-B-197 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026197,
author = {Fan, Yafu},
title = {Asymptotic Habit-Formation Curves, Debunking the 21-Day Myth and a Cue-Based Automaticity Index},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}Other chapter answers in this book
- Chapter 1 · Where Twenty-One Days Came From: Maltz's Minimum and Lally's Sixty-Six
- Chapter 3 · A Standard-Library Fitter: Grid Search, Gauss-Newton, Screening Bias and Extrapolation Error
- Chapter 4 · The Missed Day: Absence Shocks, the Dose of Compliance and the Cost of Streak Resets
- Chapter 5 · A Cue-Based Automaticity Index: Cue Entropy, Time of Day and a Repetition Clock
- Chapter 6 · Scales, Noise and the Average Curve: SRHI, SRBAI and the Trap of Right Skew
- Chapter 7 · East and West in the Classics: "By Nature Near, by Practice Far" and Aristotle on Hexis
- Chapter 8 · Closing the Loop: a Flexible Sixty-Six-Day Cycle, the Habit Curve Fitter and a Pre-Registered Plan