Loading your edition...
Back to Strata-AF™
STRATA ELECTIONSA Strata Publishing Co product

Nigeria · Projection

Presidential election 2027 · West Africa
Follow coverage
  1. Expected
  2. Announced
  3. Campaign
  4. Voting
  5. Counting
  6. Certified
  7. Transition
Stage 3 of 7 · Campaign

Projection · Nigeria · state by state

The race this week

Tinubu leads.
The 25% rule decides if it ends in round one.

We ran the election 10,000 times. Scaled to 100, Tinubu wins outright in 43, Atiku in 9, and in the other 48 nobody clears the bar — the race goes to a run-off. Obi tops the vote in 10 runs but never meets the spread rule.

① Most votes nationallyin 100 simulated elections
Tinubu 68
Atiku 22
Obi 10
② At least 25% in 25 stateseach candidate, in 100
Tinubu44
Atiku11
Obi0

An outright win needs ① and ② together. Both, in the same simulation: Tinubu 43 · Atiku 9 · Obi 0.

43likely range 37–51
Tinubu wins outright in 43 of 100

Uncertainty is wider than usual this week (×1.37). The FCT reading of the 25% rule is tested in every run. Reviewed and published by a Strata-AF™ editor.

02 · States to lead

Obi leads 13 states, Tinubu 12, Atiku 11. Kano belongs to none of them.

Colour shows who leads; strength of colour shows how firmly. Hatched states are too close to call. Play the map from 2019 to this week’s forecast to see how it moved.

Obi 13Tinubu 12Atiku 11Others 1
Osun

Forecast: leader among all candidates, central estimate.

03 · Closest states

Sokoto is a coin toss: 0.1 points between Atiku and Tinubu.

The 8 tightest states, ranked by the gap between first and second. These are where a single good week moves a state.

0%25%50%80%GAP

Dots = central forecast share of the vote. Faint bars = likely range. Tap a state to open it on the map.

04 · On the 25% line

Tinubu sits 1.5 states short of the line.

To win in the first round a candidate needs 25% of the vote in 25 of the 37 states. Every state is a tile, sorted by the candidate’s share. Solid tiles are safely over 25%, hatched tiles are on the line, empty tiles are out of reach.

Tinubu23.5 states expected · needs 259 safely over · 25 on the line · 3 out of reach
highest share25th state ▲lowest
Atiku18.5 states expected · needs 255 safely over · 25 on the line · 7 out of reach
highest share25th state ▲lowest
Obi15.4 states expected · needs 258 safely over · 13 on the line · 16 out of reach
highest share25th state ▲lowest
Which states sit on the line?
25% LINE
0%50%100% of the state’s vote

05 · Where support moved

Nobody gained this week. All three bled to the stay-at-home vote — except in Osun.

NATIONAL AVERAGE CHANGE, POINTS OF REGISTERED VOTERSAtiku−0.63Tinubu−0.40Obi−0.19Stay home+1.22
WHAT MOVED IT · TOTAL POINTS BY TYPE
Strength2.72 pts
Rallies, endorsements and defections
Exasperation0.95 pts
Supporters giving up after gaffes and attacks
Conversion0.07 pts
Voters switching camps

06 · The evidence

22 reports, each checked against its source, moved these numbers.

One block per report, stacked by the day it ran. Taller blocks moved more. Tap any block to read it.

Mon 28 SeptTue 29Wed 30Thu 1 OctFri 2Sat 3Sun 4
+3 ptsTINUBU · OSUN · SUN 4 OCT
Opposition Osun Governor Adeleke has repeatedly pledged support for Tinubu's 2027 re-election, signalling unusual cross-party backing.Conversion: winning voters from a rival. Applied to Osun only.SOURCE · THISDAY · VERIFIED BEFORE IT COUNTED

How we know

The rule, the record and the people behind every number.

What it takes to win

Most votesacross the whole country
25% in 25 of 37at least 25% of the vote in 25 of the 37 states

The most votes nationally, and at least 25% of the vote in 25 of the 37 jurisdictions (36 states and the FCT). Miss either and the leading candidates go to a run-off. Section 134(2), 1999 Constitution.

2023 by state
CandidateShareStates won≥25%
Bola Tinubu (APC)37%1229
Atiku Abubakar (PDP)29%1221
Peter Obi (LP)25%1216
Rabiu Kwankwaso (NNPP)6%11

Turnout in 2023: 27% of registered voters across the 25 states with declared figures.

How sure is this?

Median state uncertainty
×1.37
States too uncertain to call
0 of 37
Balance check
Passed
Simulations
10,000 · 200 evidence resamples

Run by sAI

sAI, Strata-AF™’s election engine, gathers campaign evidence from the news, checks each piece against its source (the supporting sentence must appear word for word, and the candidate and state must be named), and turns what passes into the weekly movement and forecast. Strata-AF™ editors set the rules sAI works to, review anything it cannot verify, and audit its calls. sAI publishes on its own only when every guardrail passes; otherwise an editor decides. Each report above says who published it.

No corrections so far. Any correction or retraction will be listed here.

Tested against 2015 → 2019

Before any forecast is trusted, the model is run as if it were the week before 2019, starting only from the 2015 results and with no campaign evidence, then scored against what happened. Structural layer only: the week before 2019, seeded from 2015, no weekly evidence.

States called right31/37pass: 30 or more
Average miss in a state’s vote share10.1 ptspass: 6 points or less
80% ranges that held the result72%pass: 70–90%
ResultNot yetneeds weekly evidence

A structural model only knows the last election. It misses realignments by design; weekly evidence is what has to close the gap, and no forecast is published here until it does and an editor signs it off.

About this data (1)
  • 2023 state figures were corrected against INEC; the national APC total from state sums is 10,929 above INEC’s declared figure, unexplained.
Sources for past results (37)