Four-Stage Pipeline

Per-seat analysis is the output of a four-stage pipeline: (1) Historical evidence base — official EC results for 2021, 2016, and 2011. (2) Field survey — voting intention and sentiment instrument, 450–675 respondents stratified by region and gender (not yet fielded). (3) Six-pillar KPI rubric — Ground organisation 30%, Historical baseline 20%, Religious & sectarian dynamics 15%, Structural factors 15%, Candidate strength 15%, Social-media signal 5%. (4) Cross-verification — independent LLM jury (Claude, GPT-4, Gemini) plus manual adjudication against EC candidate list and Pakistani press coverage.

Data Sources

• 2021 GE: Official AJK EC booklet (25 July 2021) + post-tribunal gazette. LA-15: PPP by-election 2023. LA-35: PML-N recount 2024.
• 2011 & 2016: PPP-compiled document (Amar Zeeshan Jaral, Secretary Coordination PPP AJK). 35/45 seats cross-verified.
• 2026 voter rolls: EC notification EC/S/12890-13062/2026, 22 May 2026.
• 2022 LG results: AJK local body elections by tier and district.
• 2026 candidates: Based on 2021 EC data. PTI candidates shown as IND (ex-PTI). Will be updated when EC publishes the 2026 final list.

Live Results Architecture

Live results are stored in a Supabase PostgreSQL database with real-time WebSocket subscriptions. Viewers see vote counts update the instant a data-entry team member saves a result — no polling, no page refresh. Only authenticated users can write results. All writes are logged with a timestamp and user identifier for auditability.

Independence Statement

This is an independent, data-driven analytical platform published for public reference. It is not affiliated with any political party, government body, or media organisation. All source documents are cited above. The platform publishes its methodology openly so that every figure can be audited against its inputs.