The five identified categories establish that approximately 1,00,250 votes were stolen in Mahadevapura in the 2024 general election.
Speaker / publisher: Rahul Gandhi’s 7 August 2025 presentation, published by Congress on 27 August (paraphrase)
Original date: 7 August 2025. The party’s web account was published on 27 August 2025.
Mahadevapura is an assembly segment within Bangalore Central parliamentary constituency. A general-election segment analysis must not be relabelled as a separate assembly election.
Evidence supporting the claim
Congress gives category totals for duplicate entries, invalid addresses, bulk addresses, photographs and Form 6 registrations. These add arithmetically to 1,00,250. That establishes the structure of the allegation, not independent verification of its classification. [congress-mahadevapura]
Evidence challenging the claim
An entry in a register is not itself a recorded ballot. The reviewed presentation does not bridge every step from a flagged entry to an ineligible person, a vote actually cast, a candidate benefit and a demonstrated change in the result. Nor have we verified whether categories overlap.
Relevant responses & contact record
This review does not authenticate a separate ECI rebuttal or adopt a denial as proof. We have not contacted Congress or the Commission. The assessment rests on the scope of the presented evidence, and can be revised when underlying records become available.
Insufficient evidence
The presentation supplies allegations warranting examination, but its published category counts do not independently demonstrate that 1,00,250 fraudulent ballots were cast or that the result changed. This is not a finding that every flagged registration was valid.
Limitations
- We have not independently inspected the underlying voter-level dataset, polling records, field verification or case-specific adjudication.
- Shared addresses can have legitimate explanations; they can also merit verification. Neither conclusion follows from an address count alone.
- Ballot secrecy limits inference about which candidate received an individual’s vote. An aggregate comparison with a margin is not a substitute for establishing unlawful ballots.
Read the categories as allegations
| Congress’s category | Count stated by the party |
|---|---|
| Duplicate voters | 11,965 |
| Fake or invalid addresses | 40,009 |
| Bulk registrations at one address | 10,452 |
| Missing or invalid photographs | 4,132 |
| Alleged Form 6 misuse | 33,692 |
| Arithmetic sum; not an independently deduplicated total | 1,00,250 |
Five questions between a register and a result
- Is the underlying entry authentic and from the relevant roll version?
- Does the suspected irregularity actually establish ineligibility, duplication or an administrative error?
- Did the person participate in the relevant poll, and was more than one ballot unlawfully cast?
- What evidence establishes the scale and direction of any resulting effect?
- What did the competent inquiry or court determine, under which legal test?
A large allegation deserves careful investigation. A government supporter should not dismiss a documented discrepancy merely because the complainant lost. Equally, a party cannot turn surprise at defeat into the missing evidentiary steps. Publication of an auditable methodology would make both positions easier to test.
Transparency without exposing voters
Researchers can disclose matching rules, sampling methods, aggregate errors and redacted examples without publishing a searchable dossier of private voters. Keep the raw material in the appropriate lawful review process. Public accountability does not require inviting harassment at somebody’s home.
Sources & passage references
- congress-mahadevapura / Original statement
Published account of Rahul Gandhi’s electoral-roll presentation ↗
Indian National Congress · Published 27 August 2025 · Accessed 6 October 2026
Locate: Account of press conference on 7 August 2025; paragraphs giving five categories and the figure 1,00,250
A party’s publication of its own allegations. Original presentation date and publication date differ. No underlying person-level dataset was independently audited for this article.
Publication & correction history
- 6 October 2026 · Publication
First source-managed edition. No subsequent corrections recorded.
Have a document that changes this assessment? Include the disputed passage, original source and relevant date. Submissions go to editorial review.
Submit evidence or a correction ↗