What does Short-Video Traffic Pools conclude, in numbers, about “Completion, Engagement, Replay: the Statistics of Three Rulers”?

Answer: Chapter 4 shows that a 7-second video completes 2.89 times as often as a 60-second one purely because of length, and that telling an interaction rate of 0.03 from 0.04 takes 5,301 views.

Scope of the numbers: Chapter 3 shows that a hard threshold is a ramp in the data, whose width narrows from 0.124 to 0.023 as the batch grows from 100 to 3,000 viewers; the illustrative base video breaks out with probability 0.03585, with the bottleneck at the fourth gate. Marked boxes separate published evidence, runnable Python calculations (18 programs), classical sources and the author's hypotheses.

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: 0.03, 0.04, 5301, 60

A. interaction rate 0.04 seen on n views, Wilson 95% interval
n300 0.0230 to 0.0686 width 0.0456
n3000 0.0336 to 0.0476 width 0.0141
n30000 0.0378 to 0.0423 width 0.0044
views per video to tell 0.03 from 0.045301
B. replays per view, same loop habit rho, completion from p41
L(s) completion rho0.2 rho=0.3 per completer at 0.3
70.600 0.150 0.257 0.429
150.511 0.128 0.219 0.429
300.379 0.095 0.162 0.429
600.208 0.052 0.089 0.429

Key formula

\[C(L) = (1-h)\,e^{-\lambda (L-3)},\]
Show the full program
# Chapter 4. p42: how much an early batch can say about engagement,
# and why replay rewards short videos mechanically.
# Part A: Wilson 95% interval (Wilson 1927) for an interaction rate
# measured on n views, and the views needed per video to tell 0.03
# from 0.04 with two-sided alpha 0.05 and power 0.80 (normal approx).
# Part B: a viewer who finishes the video lets it loop again with
# probability rho; expected replays per view = C * rho / (1 - rho),
# where C is the completion rate from p41. Values are illustrative.
import math

Z = 1.959964
ZB = 0.841621


def wilson(x, n):
    p = x / n
    den = 1.0 + Z * Z / n
    mid = (p + Z * Z / (2.0 * n)) / den
    half = Z * math.sqrt(p * (1.0 - p) / n + Z * Z / (4.0 * n * n)) / den
    return mid - half, mid + half


print("A. interaction rate 0.04 seen on n views, Wilson 95% interval")
for n in (300, 3000, 30000):
    lo, hi = wilson(0.04 * n, n)
    print("   n = %6d   %.4f to %.4f   width %.4f" % (n, lo, hi, hi - lo))
p1, p2 = 0.03, 0.04
pb = (p1 + p2) / 2.0
need = (Z * math.sqrt(2 * pb * (1 - pb))
        + ZB * math.sqrt(p1 * (1 - p1) + p2 * (1 - p2))) ** 2 \
    / (p2 - p1) ** 2
print("   views per video to tell 0.03 from 0.04: %d" % math.ceil(need))
print("B. replays per view, same loop habit rho, completion from p41")
print("   L(s)  completion  rho=0.2  rho=0.3  per completer at 0.3")
for L, C in ((7, 0.600), (15, 0.511), (30, 0.379), (60, 0.208)):
    r2 = C * 0.2 / 0.8
    r3 = C * 0.3 / 0.7
    print("   %4d  %10.3f  %7.3f  %7.3f  %8.3f" % (L, C, r2, r3, r3 / C))

Single source and citation

Fan, Y. (2026). Cold-Start Dynamics of Short-Video Traffic Pools: Step Thresholds in Completion, Engagement and Replay and a Zero-Ad-Spend Breakout Test Matrix (Version 1.0.0), Chapter 4: Completion, Engagement, Replay: the Statistics of Three Rulers. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-273 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026273,
  author    = {Fan, Yafu},
  title     = {Cold-Start Dynamics of Short-Video Traffic Pools: Step Thresholds in Completion, Engagement and Replay and a Zero-Ad-Spend Breakout Test Matrix},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book