For law firmsIReV forensicsConfidentiality & conflictsEvidence & investigationsDiscuss a petition ↗
For election litigation teams

Election Petition Analytics.
Form EC8A Forensics.

Evidence Your Team Can Use.

MundX turns Form EC8As into court-ready evidence for election petitions. Through detailed analysis and document forensics, we reconcile the figures, examine the sheets for alterations and inconsistencies, and show exactly which polling units are in issue and whether they change the result. Every finding traces back to its source. You get an evidence schedule, technical briefing and, where needed, a witness statement on oath, within 48 hours to 7 days. We act for petitioners and respondents, and all work is strictly confidential under NDA.

Inspect real polling-unit findings ↗
Confidentiality & conflicts

Your case stays confidential.

Conflicts disclosed

We check for conflicts and disclose relevant involvement before accepting instructions.

NDA first

We execute an NDA before you share confidential case files or litigation strategy.

Controlled access

Files move securely, with access limited to the agreed team and handling terms set with counsel.

Start with party details and a high-level brief.

For petitioners and respondents

Support for the side you represent.

Petitioner counsel

Build the petition record.

Move from a broad concern about the election to named units, specific discrepancies and a clear account of what the figures support.

  • Identify discrepancies that support specific petition particulars.
  • Prioritise disputed units by candidate-specific numerical effect and record quality.
  • Identify records to seek from INEC and questions requiring firsthand evidence.
Respondent counsel

Test the opposing case.

Examine whether the pleaded units, documents and calculations support the allegation and the numerical conclusion advanced.

  • Reproduce the opposing tally and isolate calculation or transcription errors.
  • Test allegations against the cited documents, coverage and assumptions.
  • Assess whether the proposed treatment changes a candidate’s score or the margin.
Section 138(1) · Electoral Act 2026

Analysis aligned with the grounds of petition.

Section 138(1)(a)

Corrupt practices or non-compliance

“invalid by reason of corrupt practices or non-compliance”

Identify the affected polling units, examine documentary inconsistencies and organise the records and questions counsel needs to prove them.

Section 138(1)(b)

Majority of lawful votes

“not duly elected by majority of lawful votes”

Reconcile polling-unit figures, test disputed adjustments and calculate their effect on candidate totals and the margin.

Excerpts from the two grounds in Section 138(1) of the Electoral Act 2026 ↗. Section 139(1): non-compliance does not invalidate an election conducted substantially in accordance with the Act. Read Section 139(1) ↗

Digital forensics for IReV Form EC8As

The figures are one question. The file behind them is another.

Examine the digital files published on the INEC Result Viewing (IReV) portal to test how the forms were captured, processed and published.

From Form EC8As to a reviewable evidence pack

  1. Step 1Source recordsForm EC8As, BVAS accreditation records and collation forms.
  2. Step 2Check & reconcileIdentity, figures, metadata and available versions.
  3. Step 3Explain the findingCalculation, assumptions, possible effect and evidence gaps.
  4. Step 4Brief counselSource-linked schedules, technical findings and focused record requests.
What the forensic instruction covers

Capture metadata & timelines

Extract recoverable camera, software, time and time-zone fields from the Form EC8A images embedded in IReV files.

Document identity & replacements

Compare printed polling-unit codes, form serials, portal labels and successive images. Identify mismatches and replacements.

Cross-image comparisons

Examine shared backgrounds, framing and device-model profiles alongside file hashes. Distinguish duplicates from distinct captures.

IReV source preservation & explanation

Retain Form EC8A files retrieved from IReV or supplied for comparison, record their identifiers and hashes.

We identify questions that require original-device examination, authenticated logs or official inspection.

Start before the filing clock becomes the constraint

Your case has a deadline. The record needs a head start.

Begin the geography, source inventory and record checks before declaration where possible, so the 21-day window is spent on findings, not set-up.

The petition clock

Before declarationGeography, source inventory, record checks
  1. Day 0Declaration
  2. Day 7
  3. Day 14
  4. Day 21Petition filed with witness statements & documents

Illustrative. Counsel should confirm the applicable computation of time for the specific matter.

Historical investigations and 2023 judgments may refer to the Electoral Act 2022. Current service references use the Electoral Act 2026.

What your law firm receives

Four outputs for your litigation team.

Evidence schedule

Named polling units, source values and discrepancies mapped to counsel’s issues, with outstanding evidence flagged.

Reconciliation workbook

Candidate totals, coverage, missing units and numerical scenarios, with reproducible calculations and assumptions.

Forensic & source pack

Metadata exports, identity and version comparisons, a source index and file hashes, with methodology notes.

Counsel briefing & witness support

A technical conference to examine the findings. Where required, a witness statement on oath from the analyst.

Real records from MundX investigations

Findings you can examine.

Examples span different elections and illustrate distinct analytical questions. They are not a combined dataset.

01 / Digital document forensics · Imo, 2023

Before polls opened. An iPhone recorded as the capture device.

24distinct Form EC8A images
8,081stated valid votes in the cohort

The embedded metadata raises two questions:

Why do the files record capture times before voting began?
The images record times between 05:04:19 and 05:34:17 WAT on 15 April 2023—roughly three hours before the scheduled 8:30am start of polling.

Why do they identify an iPhone rather than BVAS?
All 24 images identify Apple iPhone 13 / iOS 16.4.1 in their embedded metadata, rather than a BVAS device.

For counsel: establish whether the device clock was accurate, when the result sheets were completed, and how these iPhone images entered the official result-publication process. Metadata alone does not establish the actual capture time or the full handling history.

Read the forensic investigation ↗ · 24-Form EC8A dataset ↓
Supplementary Form EC8A for Mbaitoli unit 16/11/09/011
Example: 16/11/09/011 · Eke Ubaha Market Square · 321 valid votes. Embedded original-time field: 05:31:20 WAT, 15 April 2023.
Original IReV PDF ↗ · Metadata export ↓

Embedded capture times, all 24 images · 15 April 2023 (WAT)

Each mark is one Form EC8A. Device field on every image: Apple iPhone 13 · iOS 16.4.1.

05:04:19 · Ikeduru23 images · 05:24:07–05:34:1716/08/11/005: 05:04:19 WAT16/11/11/001: 05:24:07 WAT16/11/02/007: 05:25:06 WAT16/11/02/003: 05:25:33 WAT16/11/02/001: 05:25:58 WAT16/11/02/008: 05:26:17 WAT16/11/02/006: 05:26:34 WAT16/11/02/005: 05:27:01 WAT16/11/02/004: 05:27:16 WAT16/11/02/002: 05:27:33 WAT16/11/02/021: 05:28:15 WAT16/11/02/015: 05:28:29 WAT16/11/02/013: 05:28:46 WAT16/11/02/012: 05:29:01 WAT16/11/02/011: 05:29:21 WAT16/11/02/010: 05:29:34 WAT16/11/02/009: 05:29:49 WAT16/11/01/016: 05:30:21 WAT16/11/01/009: 05:30:39 WAT16/11/09/011: 05:31:20 WAT16/11/09/016: 05:31:34 WAT16/11/09/012: 05:31:49 WAT16/11/06/008: 05:34:02 WAT16/11/06/013: 05:34:17 WAT05:0405:1005:1505:2005:2505:3005:35▲ 16/11/09/011 · 05:31:20

Source: embedded original-time fields in the IReV-published files (24-Form EC8A dataset). Times as recorded by the device; clock settings are part of what counsel should test.

02 / Result reconciliation · Edo, 2024

A 35,631-vote difference between two margins.

MundX’s count of available Form EC8As covers 4,460 Form EC8As: APC 269,595; PDP 260,833—an 8,762 lead. Declared totals of APC 291,667 and PDP 247,274 give a 44,393 lead. The difference between those margins is 35,631.

APC lead over PDP, two ways

45,00022,5008,762+35,63144,393Lead in availableForm EC8AsDifference toexplainDeclared lead

4,460 forms

Where the difference sits

PartyForm countDeclaredChange
APC269,595291,667+22,072
PDP260,833247,274−13,559

+22,072 for APC and −13,559 for PDP together make the 35,631.

75 Form EC8As with votes above accreditation as written. A separate question from the margin reconciliation.

For counsel: inspect named Form EC8As, reconcile coverage and corrections, and determine which records explain the return.

Inspect a reviewed Form EC8A ↗
03 / Identity & replacement analysis · By-elections, 2026

One form identity.
Different portal records.

Across 1,639 polling-unit results in five contests, the investigation identifies conflicting figures and amended images.

Gombe: compare the two versions

Image listed under 15/06/08/002Image listed under 15/06/08/020
Printed on the formListed 002Listed 020Result
Polling-unit code020020Same
Form serial05140514Same
Valid votes2613Changed

Udu: four totals on one form

Unit 10/18/09/042 · Kano Rd O/S DSC, Orhuwhurun

Accreditation17
Valid total (top)17
Party sum420
Valid total (bottom)430

Party sum = APC 415 + APM 3 + PDP 2. It is 403 above accreditation and 10 below the bottom total. Establish which figures official collation records substantiate.

Open original Form EC8A ↗
Gombe: compare the two versions

Images listed under 15/06/08/002 and 020 both print unit 020 and serial 0514. The stated valid total changes from 26 to 13. Resolve the portal-to-form mismatch and establish the authoritative record.

Gombe image listed under unit 002, printing unit 020 and serial 0514, valid votes 26Gombe image listed under unit 020, same serial 0514, amended valid votes 13
Udu: inspect conflicting totals

Unit 10/18/09/042 · Kano Rd O/S DSC, Orhuwhurun. Written accreditation is 17; APC 415, APM 3 and PDP 2 total 420. Valid totals are 17 at the top and 430 at the bottom. The party sum is 403 above accreditation and 10 below the bottom total. Establish which figures official collation records substantiate.

Udu Form EC8A showing conflicting accreditation and valid-vote totals

For counsel: identify the document and version under challenge, reconcile the amendments and prepare focused requests for the underlying records.

Full investigation data ↓
04 / Coverage & sensitivity analysis · Rivers, 2023

When the available count and the return point different ways.

The available IReV count places LP ahead by 13,043 in Rivers East Senate at 81.36% coverage, and by 30,778 in Obio/Akpor House at 96.37% coverage. INEC returned PDP candidates in both contests.

Record coverage vs. the available-count lead

Rivers East Senate

LP +13,043 in available count

INEC return: PDP
Obio/Akpor House

LP +30,778 in available count

INEC return: PDP

The hatched share is why sensitivity testing has to precede a different-winner submission.

For counsel: quantify the missing-record problem and test assumptions before advancing a result theory. Prioritise the 653 unusable Senate units for record recovery and sensitivity testing.

653-unit reconstruction dataset ↓ · Ward summaries ↓
05 / Forensic audit at scale · Presidential election, 2023

164,659 accessible records inspected.

Across IReV-published result records, the audit found EMP2920/BVAS model metadata in 52,947 records, another identifiable model in 50,452, and no recoverable model in 61,260.

Device model found in the file metadata

Denominator: 164,659 accessible IReV result records. Shares rounded to one decimal.

For counsel: use a documented population, denominators and a unit-level source path to target technical inquiries.

Read the device-metadata audit ↗ · State summary ↓

Methodology: Findings retain their source, calculation and assumptions. Discrepancies, reconstructed totals and metadata patterns require corroboration; they do not alone establish invalid votes, fraud, a unique device or a different lawful winner. Counsel assesses the evidence and applicable law; the court determines admissibility and the merits.

How to instruct MundX

Start with a focused assessment.

Questions from counsel

Ready for the questions behind the instruction.

What data sources can you work with?

We can work with data from any source: IReV Form EC8As, BVAS machine accreditation records, certified true copies and collation records.

Can you supply a witness statement on oath in time to file with the petition?

Yes, on the timetable agreed at instruction. We coordinate the statement, supporting schedules and exhibits with counsel.

What if the analysis does not support our proposed case?

We report that finding. Counsel receives the supporting calculations, the limitations and any unresolved questions.

Discuss your matter with MundX.

Request a focused assessment of the support your team needs.

[email protected]

MundX · Legal support

Let’s scope your case support.