Most exit polls say Vijay’s TVK won between 10 and 24 seats. One says 98 to 120. Counting on May 4 will tell us which pollster read the new Tamil Nadu — and which one was just averaging assumptions.
That’s the story buried inside the consensus. After Tamil Nadu and West Bengal voted last week, six major pollsters released numbers on April 29. On the surface, they agreed: DMK holds Tamil Nadu, West Bengal is too close to call. The disagreements underneath are where the real news is.
The TVK Outlier Nobody Can Explain
Matrize gives DMK 122–132 seats. Chanakya Strategies has DMK at 145–160. Most pollsters then hand Vijay’s TVK between 10 and 24 seats — a debut, not a takeover. Then Axis My India breaks ranks. It puts TVK at 98–120 seats. The majority threshold is 118.
That isn’t a difference of degree. It’s a different election. Vijay contested two seats himself, and one of these pollsters is reading vote-bank movement the others missed entirely. Tamil Nadu’s turnout hit 85.1% — the highest in the state’s assembly history. Both interpretations live inside that number.
West Bengal Can’t Decide What It Saw
Phase 1 turnout crossed 93%. P-Marq gives BJP 150–175 seats and TMC 118–138. Matrize lands BJP at 146–161. People’s Pulse hands TMC 177–187 and writes off the BJP surge entirely.
The polls aren’t reading the same state. They’re reading the same anti-incumbency signal — 15 years of Mamata, the SIR controversy, women’s safety, jobs — and reaching opposite conclusions about whether it broke or held. The 2021 turnout was 85.2%. This time voters showed up in larger numbers, and pollsters can’t agree whether that means change or reaffirmation.
What May 4 Actually Tests
Exit polls aren’t predictions. They’re reports of what voters told strangers with clipboards three days ago. India’s 2026 political story won’t be set by who wins on Tuesday. It’ll be set by which pollster was actually reading voters — and which spent the last week filling in blanks with old maps.