What does Citation Weight in Generative Search conclude, in numbers, about “Measuring Visibility: Why Keyword Stuffing Fails and Why Sources, Figures and Quotations Help”?
Answer: Chapter 1 rebuilds the paper's table of effects and uses a proportional-share model to show that a rewrite adopted alone buys only a 0.3027 gain in share and nothing once all adopt.
Scope of the numbers: Its parameters are illustrative; all calculations are generated by its programs. It is not individual business advice.
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: 0.3027
| sources per answer N = 5, weight multiplier g = 1.4093 | |||
| adopters | adopter shr | gain vs 1/N | non-adopter |
| 0 | - | - | 0.2000 |
| 1 | 0.2605 | +0.3027 | 0.1849 |
| 2 | 0.2422 | +0.2110 | 0.1719 |
| 3 | 0.2263 | +0.1314 | 0.1606 |
| 4 | 0.2123 | +0.0617 | 0.1507 |
| 5 | 0.2000 | +0.0000 | - |
| lone adopter: weight +0.4093 buys share +0.3027 | |||
| all adopt: every share back to 0.2000, gain 0 |
Show the full program
# Chapter 1. p13: why a rewrite gain shrinks when competitors adopt it.
# Model: an answer splits a fixed attention budget among N sources in
# proportion to weights. A rewrite multiplies a source's weight by g.
# With a adopters among N, an adopter's share is g / (a g + N - a);
# its gain over the plain share 1 / N falls to zero when all adopt.
# g = 1.4093 is the best PAWC ratio of p11 (quotation addition); the
# proportional-share model is ours, an illustration of the congestion
# reported by Puerto et al. (2025), not their data.
from fractions import Fraction as F
N = 5
G = F(272, 193) # 27.2 / 19.3 from Aggarwal et al. Table 1
print("sources per answer N = %d, weight multiplier g = %.4f" % (
N, float(G)))
print("%9s %13s %13s %13s" % ("adopters", "adopter shr", "gain vs 1/N",
"non-adopter"))
for a in range(0, N + 1):
total = a * G + (N - a)
ad = G / total if a else None
non = 1 / total if a < N else None
print("%9d %13s %13s %13s" % (
a, "-" if ad is None else "%.4f" % float(ad),
"-" if ad is None else "%+.4f" % float(ad * N - 1),
"-" if non is None else "%.4f" % float(non)))
lone = G / (G + N - 1) * N - 1
print("lone adopter: weight +%.4f buys share +%.4f" % (
float(G - 1), float(lone)))
print("all adopt: every share back to %.4f, gain 0" % (1 / N))Single source and citation
Fan, Y. (2026). The Geometry of Citation Weight in Generative Search: Retrieval and Reranking Features and the Paragraph Topology of Highly Cited Bilingual Answer Pages (Version 1.0.0), Chapter 1: Measuring Visibility: Why Keyword Stuffing Fails and Why Sources, Figures and Quotations Help. 5D Library. DOI pending
DOI pending · monograph page goes live after publication · 5DL-B-260 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026260,
author = {Fan, Yafu},
title = {The Geometry of Citation Weight in Generative Search: Retrieval and Reranking Features and the Paragraph Topology of Highly Cited Bilingual Answer Pages},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}Other chapter answers in this book
- Chapter 2 · The Geometry of Hybrid Scoring: BM25 Saturation, Cosine Direction and Rank Fusion
- Chapter 3 · Paragraph Length and the Point of the Cut: Chunking, Truncation and an Interior Optimum
- Chapter 4 · The Discount by Citation Position: Cumulated Gain, Position-Adjusted Word Count and the Sag in the Middle
- Chapter 5 · Verifiability: Why Sources Raise the Chance of Citation, and the Trap of Inferring Features from Outside
- Chapter 6 · The Paragraph Topology of a Bilingual Direct-Answer Page: Self-Contained Paragraphs, One Set of Numbers on Four Surfaces, and Compliant Structured Data
- Chapter 7 · Classical Sources East and West: the Three Gauges of the Mozi and Garfield's Citation Index
- Chapter 8 · Closing the Loop: an A/B Test of Answer-Paragraph Templates, the Citability Scorer and a Pre-Registered Test