2028 presidency — Jon Ossoff

Conditional on 2026 GA Senate retention (EHIQ 62% to retain per senate-ga-ossoff-2026 entry). If he wins big in 2026, becomes credible 2028 dark horse. Polymarket 3.8% now ALIGNED with EHIQ post-SEM-build. SEM v1.2 D-nomination = 16.85% (poll baseline 2-3% Ossoff share × long-lead compression x0.6 + withdrawal multiplier 0.40 reflecting bimodal Senate-vs-presidency path conditional on 2026 outcome). Apples-to-apples presidency: 16.85% × ~35% general electability with first-term-Senator-jump-discount ≈ 5.9%. 50% damping from prior 2.5% baseline: 2.5 + 0.5*(5.9-2.5) ≈ 4.2% → 4%. Note: separate `sem_ga_senate_ossoff_2026.json` handles the 2026 GA Senate retention question (ensemble 82.36%).

• •Path entirely conditional on 2026 Senate reelection. Margin matters: a comfortable win opens the lane; a narrow win or loss eliminates it.
• •Swing-state appeal + youth + 2020/2021 runoff brand.
• •Not yet EHIQ register-scored at 2028-presidency level (2026 GA Senate corpus exists — DIFFERENT race-id).
• •PROBABILITY REVISION TIMESTAMPED 2026-05-22 (PART 2): EHIQ 2.5% → 4% after SEM v1.2 build. New SEM file `sem_us_president_2028_ossoff.json` / `.prov.json` (RACE_ID = us-president-2028-dem-nominee; do NOT conflate with separate ga-senate-2026-general SEM). SEM v1.2 D-nomination = 16.85% (raw P 13.48% → long-lead compression x0.6 → 28.1% → withdrawal multiplier 0.40 for bimodal Senate-vs-presidency path → 16.85%). Apples-to-apples ≈ 16.85% × 35% gen electability ≈ 5.9%. 50% damping from prior 2.5% baseline → 4%. INSUFFICIENT_POLLING flag. Note: there IS a 2026-GA-Senate TIQ corpus for Ossoff but it is a DIFFERENT race-id and the ensemble script may attempt cross-race lastname merge — verify no contamination after ensemble run.

Read a full sample of the paid research →