Fix the expected units
The INEC polling-unit register supplies stable territory or state, local, ward and unit codes. Kosofe contains 899 inventory units in this snapshot, which prevents source rows from silently changing the denominator.
Local evidence record · Declaration unavailable
Kosofe 2023 presidential election evidence
Mundx's selected polling-unit observations place LP first in Kosofe's four-party aggregation with 45,395 votes, ahead of APC by 9,304. This is not an official LGA declaration. The current source register contains no separately sourced Kosofe collation record, so the declared-result field remains unavailable.
Evidence status: complete inventory representation for the four selected parties does not make these figures an EC8C, EC8D or other declared collation result.
Kosofe PU layer
Four-party polling-unit aggregation
The selected records total 86,965 votes across APC, LP, PDP and NNPP. LP accounts for 52.2% of that four-party sum; APC accounts for 41.5%.
| Party | Selected PU votes | Four-party share |
|---|---|---|
| LP | 45,395 | 52.2% |
| APC | 36,091 | 41.5% |
| PDP | 4,263 | 4.9% |
| NNPP | 1,216 | 1.4% |
| Four-party total | 86,965 | 100.0% |
No declared comparison is shown. Mundx will add a declared column only when the source register contains a traceable LGA collation record with its authority, form level, retrieval date and confidence.
Ward evidence
How the four-party pattern varies within Kosofe
APC leads 3 and LP leads 7 of the 10 ward rollups. Every row is calculated from selected PU observations and keeps the representation count beside the vote totals.
| Ward | PU leader | APC | LP | PDP | NNPP | Four-party | PUs represented |
|---|---|---|---|---|---|---|---|
| Agboyi I | APC | 507 | 89 | 83 | 2 | 681 | 6 / 6 |
| Agboyi Ii | APC | 366 | 36 | 46 | 1 | 449 | 5 / 5 |
| Anthony/Ajao Estate/Mende/Maryland | LP | 1,670 | 4,302 | 212 | 49 | 6,233 | 51 / 51 |
| Ifako/Soluyi | LP | 2,622 | 5,935 | 259 | 40 | 8,856 | 79 / 79 |
| Ikosi Ketu/Mile 12/Agiliti/Maidan | APC | 8,608 | 5,673 | 798 | 246 | 15,325 | 192 / 192 |
| Isheri-Olowo-Ira/Shangisha/Magodo Phase I & Ii | LP | 4,841 | 6,607 | 632 | 387 | 12,467 | 124 / 124 |
| Ketu/Alapere/Agidi/Orisigun/Kosofe/Ajelogo/Akanimodo | LP | 7,409 | 7,580 | 989 | 228 | 16,206 | 179 / 179 |
| Ojota/Ogudu | LP | 4,289 | 6,739 | 447 | 110 | 11,585 | 115 / 115 |
| Owode Onirin/Ajegunle/Odo-Ogun | LP | 2,718 | 3,423 | 323 | 58 | 6,522 | 59 / 59 |
| Oworonshoki | LP | 3,061 | 5,011 | 474 | 95 | 8,641 | 89 / 89 |
Largest ward rollups
Five local records in context
These notes restate the largest ward aggregates in prose so their leader, margin and representation remain readable outside the table.
- Ketu/Alapere/Agidi/Orisigun/Kosofe/Ajelogo/Akanimodo: LP leads with 7,580 of 16,206 four-party votes, a margin of 171 over APC. The rollup represents 179 of 179 inventory units.
- Ikosi Ketu/Mile 12/Agiliti/Maidan: APC leads with 8,608 of 15,325 four-party votes, a margin of 2,935 over LP. The rollup represents 192 of 192 inventory units.
- Isheri-Olowo-Ira/Shangisha/Magodo Phase I & Ii: LP leads with 6,607 of 12,467 four-party votes, a margin of 1,766 over APC. The rollup represents 124 of 124 inventory units.
- Ojota/Ogudu: LP leads with 6,739 of 11,585 four-party votes, a margin of 2,450 over APC. The rollup represents 115 of 115 inventory units.
- Ifako/Soluyi: LP leads with 5,935 of 8,856 four-party votes, a margin of 3,313 over APC. The rollup represents 79 of 79 inventory units.
Evidence beneath the wards
Largest represented polling-unit rollups
The ten rows below are ordered by four-party vote count. This is not a turnout ranking and does not imply official validation. The source label identifies which selected observation supplied the four party fields.
| Polling unit | Ward | PU leader | Four-party votes | Selected source |
|---|---|---|---|---|
24-13-07-012Junction Of Alh. Basser Shittu Avenue/Quadri Adebiji St. | Isheri-Olowo-Ira/Shangisha/Magodo Phase I & Ii | LP | 555 | CCIJ |
24-13-02-064Inside Atunrase Estate Phase Ii | Ifako/Soluyi | LP | 517 | CCIJ |
24-13-07-014Open Space Beside Magodo/Shangisha Police Post | Isheri-Olowo-Ira/Shangisha/Magodo Phase I & Ii | LP | 406 | CCIJ |
24-13-04-044Ramat Crescent/Ojota Road Junction | Ojota/Ogudu | LP | 386 | CCIJ |
24-13-07-051Entrance Of Unilag Estate | Isheri-Olowo-Ira/Shangisha/Magodo Phase I & Ii | NNPP | 382 | CCIJ |
24-13-10-006Irawo Road, In Front Of House No. 10 | Owode Onirin/Ajegunle/Odo-Ogun | LP | 365 | CCIJ |
24-13-04-064Omotayo Banwo St. In Front Of House No. 1 | Ojota/Ogudu | LP | 359 | CCIJ |
24-13-07-016Boy'S Approve School Compound, Isheri St. | Isheri-Olowo-Ira/Shangisha/Magodo Phase I & Ii | LP | 358 | CCIJ |
24-13-07-043Isheri Grammar Sch., Isheri | Isheri-Olowo-Ira/Shangisha/Magodo Phase I & Ii | LP | 355 | CCIJ |
24-13-05-002Comprehensive High Sch., Agidi | Ketu/Alapere/Agidi/Orisigun/Kosofe/Ajelogo/Akanimodo | LP | 346 | CCIJ |
How this record was built
From inventory to local aggregation
Apply one policy
Complete CCIJ observations take precedence, followed by complete Mark Essien fallbacks. Kosofe's selected records break down as CCIJ 880, MARK 19. The policy is deterministic and preserves the original source label.
Sum without declaring
Party values are summed from PU to ward and LGA. Missing observations stay visible instead of being estimated. In Kosofe, 899 of 899 units are represented, while the scope still excludes fourteen parties.
What this rollout preserves. The page is designed to survive a later declared-result addition without changing its URL. A future collation record would appear as a separately sourced layer beside this aggregation; it would not overwrite the PU evidence, backfill a missing total, or turn the calculated ward table into an official document. Every future update should retain the review date and show which layer changed, so readers can distinguish new source evidence from a presentation revision. Keeping that contract stable now is more important than generating hundreds of local pages at once.
Questions this page answers
Kosofe evidence notes
Did LP officially win Kosofe?
This page does not make that claim. It says LP leads Mundx's four-party aggregation of selected polling-unit observations. A statement about the declared LGA result requires a sourced collation record at the appropriate result layer.
What does representation mean?
It reports how many inventory units have a selected observation containing all four tracked party values. It does not establish sheet authenticity, capture all eighteen parties, or reproduce every correction made during collation.
Why preserve an unavailable field?
An explicit unavailable value is safer than promoting a calculated PU sum into an official result. It also makes future source work concrete: add the declared layer only with a form level, authority URL, retrieval date and documented confidence.
How is this different from the Kosofe place page?
The permanent Kosofe place record describes geography across cycles. This contest page is tied to the 2023 presidential election, its evidence layers and its source limitations. The URLs serve different questions and should not duplicate one another.
Provenance
What is official, and what Mundx calculated
The current source register does not contain a declared LGA collation record for this geography. Mundx therefore leaves the declared layer empty instead of inferring one from polling units.
Mundx's second layer is a four-party aggregation of selected polling-unit observations. It applies policy all-complete-ccij-first-v1: complete CCIJ records take precedence, then complete Mark Essien records fill remaining units.
- INEC 2023 presidential IReV record.
- State/FCT collation reproduction.
- 176,409 represented units from an inventory of 176,846; 437 remain unresolved nationally.
Canonical PU policy all-complete-ccij-first-v1 · CCIJ commit e38e3bc9a22a
Reuse boundary. Mundx Analytics confirms permission to publish and redistribute the source-derived records in this release. No standard open-data licence or third-party bulk-reuse terms are specified.
Published by Mundx Analytics · reviewed by Mundx Analytics Data Team on 22 September 2026 · official sources retrieved 2026-08-20.