The evidence is primary, the number is derived
Every figure here is read off a photographed EC8A carrying a geotag, a timestamp and an identified agent. Accredited agents feed the tally; open citizen reports are a separate stream that can corroborate and raise incidents but never enters the count. The platform reports what was collected — it does not declare a winner.
Coverage
Stated before any result — a dashboard that hides its gaps is not evidence151 forms were captured with no network and synced when signal returned, preserving the original timestamp and GPS. In Baruten, Kaiama, Edu and Pategi that is the normal case, not the exception — a system that requires connectivity at the moment of capture simply does not collect those wards.
Coverage is against 2,887 polling units — the state total after INEC converted 1,015 voting points in May 2021. Only 1,872 of those are enumerated to ward level in the directory we hold, so 1,015 units are counted in the denominator but cannot yet be placed on the ward map below.
Against the enumerated subset alone the figure would read 52.6% rather than 34.1%. The flattering number is not the one displayed. Obtaining the current directory from INEC Kwara closes the gap.
Where the returns are
Row width = LGA votes cast 2023 · cell width = polling unitsCollected tally
Tier one only — accredited agents- Votes counted
- 242,003
- Turnout (reporting)
- 37.1%
- Rejected
- 5,688
- Disputed
- 3
Turnout is against the registered voters on the 985 forms actually received, not the state register — quoting it against the full roll while a third of units are outstanding would be misleading. 3 units have a second submission that disagrees; those are published as disputed rather than silently resolved.
Incident desks
32 openRouting is deterministic by type and severity, so nobody has to decide under pressure. Every incident closes visibly and the reporter is notified — a report that vanishes teaches people to stop reporting, and the sensor network is gone by mid-morning.
Incident queue
Most recent first| Type | Severity | Location | Tier | Desk | Status | Last action |
|---|---|---|---|---|---|---|
Late start Officials and materials arrived well after the scheduled hour. | high | Kpada I Pategi | agent | Operations | Unfounded | Checked against a second source; not substantiated. 18:43 · operations desk |
Late start Officials and materials arrived well after the scheduled hour. | high | Apado Ilorin East | agent | Operations | Actioned | Escalated to the field team. 18:05 · operations desk |
BVAS failure Machine failing to authenticate; queue stalled. | medium | Osi I Ekiti | agent | Operations | Resolved | Closed. Outcome recorded and reporter notified. 18:33 · operations desk |
Obstruction of agent Agent refused entry to the polling unit by officials. | medium | Okuta Baruten | agent | Legal | Open | Reported. 16:54 · Polling agent |
Material failure Result sheets exhausted before close of poll. | low | Patigi IV Pategi | citizen | Operations | Open | Reported. 16:52 · Citizen report |
Obstruction of agent Agent refused entry to the polling unit by officials. | low | Babadudu Moro | agent | Legal | Acknowledged | Picked up by the desk. 17:15 · legal desk |
Intimidation Group of men stationed at the entrance questioning voters. | low | Oke Aba Isin | citizen | Legal | Actioned | Escalated to the field team. 18:03 · legal desk |
Inducement Cash being offered beyond the perimeter fence. | medium | Shawo Central Offa | agent | Legal | Resolved | Closed. Outcome recorded and reporter notified. 18:12 · legal desk |
Intimidation Group of men stationed at the entrance questioning voters. | high | Kpada I Pategi | agent | Legal | Acknowledged | Picked up by the desk. 16:43 · legal desk |
BVAS failure Machine failing to authenticate; queue stalled. | high | Ojomu Central II Offa | agent | Operations | Resolved | Closed. Outcome recorded and reporter notified. 17:26 · operations desk |
Late start Officials and materials arrived well after the scheduled hour. | low | Efue/Berikodo Asa | agent | Operations | Resolved | Closed. Outcome recorded and reporter notified. 17:37 · operations desk |
Obstruction of agent Agent refused entry to the polling unit by officials. | low | Osi I Ekiti | agent | Legal | Open | Reported. 16:02 · Polling agent |
BVAS failure Machine failing to authenticate; queue stalled. | low | Omu-Aran I (aran) Irepodun | agent | Operations | Actioned | Escalated to the field team. 16:38 · operations desk |
Late start Officials and materials arrived well after the scheduled hour. | medium | Marafa/Pepele Ilorin East | citizen | Operations | Resolved | Closed. Outcome recorded and reporter notified. 17:54 · operations desk |
BVAS failure Machine failing to authenticate; queue stalled. | medium | Yowere II/Okeweru Asa | agent | Operations | Resolved | Closed. Outcome recorded and reporter notified. 17:33 · operations desk |
Late start Officials and materials arrived well after the scheduled hour. | medium | Owu Isin Isin | agent | Operations | Unfounded | Checked against a second source; not substantiated. 16:07 · operations desk |
Citizen reports
195 surfaced of 260Queue moving steadily since morning.
Voting orderly, no issues observed.
Voting orderly, no issues observed.
Queue moving steadily since morning.
Queue moving steadily since morning.
Over a hundred waiting, one machine.
Open reports need 2 independent corroborations before they surface, and they never enter the tally under any circumstances. Crowdsourcing is a sensor network and an incident feed — the moment it touches the count, the count stops being evidence.
Agent readiness
1872 polling unit positionsAccreditation is what separates a tier-one submission from a tier-two report. An agent without an accreditation reference cannot file a result, only an observation.