Benford Bench

Welcome to the next generation of Benford Bench — an online data assessment & publishing tool for research in fraud and disorder detection using Benford's Law digit analysis.

Featured

Forensic Analysis of Ballot-Scan Events versus Official Device Tallies

2026-09-26

Refactored from 2020-2024-2026.pdf (pdfTeX, 19 July 2026). The file is a packet of two public analytical reports by Jason Page, Benford Bench (https://BenfordBench.org). The opening fax cover sheet is omitted, including the personal phone number, email address, and street address printed on it. The first page of the…

Read post

Benford Analysis 2020 vs 2024 vs 2026

2026-09-26

Source: Benford First-Digit Analysis of Inter-Transaction Time Deltas > 9 Seconds, Jason Page, Benford Bench, 19 July 2026. Public analytical report on Chicago ImageCast precinct logs for 3 November 2020, 5 November 2024, and 17 March 2026. File: benford-gt9-2020-2024-2026.pdf. Title page: 2020-26.png. Companion to…

Read post

2020 vs 2024 Chicago: Assessment of In-Election Machine Logs Ballot Scan Count Checks and Benford’s Law Analysis

2026-07-19

Significant anomalies are identified under both analyses in both election cycles, but the two elections present qualitatively different signatures. The 2020 logs show temporal irregularities consistent with log-level manipulation --- batch insertion, timestamp regularization, or post-hoc log editing --- reflected in…

Read post

Upload CSV for Digit Analysis

Reference Documents

0) Introductory article: one.pdf
1) Benford's Law with Time Stamps (catching cheaters): kottke.pdf
2) Benford's Law in legal history: lawhistory.pdf

CSV File:	Required. Must be a .csv file.
Column #:	Which column contains the numbers to analyze (1 = first column).
	Skip first row (header row) Check this if the first row contains column names, not data.
Analysis Type:	First digit is most common for fraud detection. Second digit adds confidence.
Delimiter:	Automatic detection reads the first line and picks comma, semicolon, or tab.
FOIA PDF:	Optional. Attach a supporting FOIA or source document (PDF).

Source of Data:	Where did this dataset come from?
Subject:	Brief description of what this data represents.
Investigative Reason:
Hypothesis:

Name:
Position/Role:
Contact:	Stored in the report only; not published publicly.