base rate:
Historical base rate of losing major-party nominees running again in the next cycle is approximately 15% (2/13 cases). This is the strongest anchor. However, structural conditions are meaningfully different: Harris entered 2024 with only 107 days, her loss is attributable partly to Biden's withdrawal timing, she's relatively young (61), and there's no incumbent Democrat.
evidence updates:
Harris's April 2026 'I might, I'm thinking about it' statement is the single strongest piece of evidence, pushing up by ~15-20% from base rate. Declining the CA governor's race pushes up by ~5-8% (clearly preserving presidential option). Book tour in early primary states pushes up by ~3-5%. Early polling lead pushes up by ~5%. Crowded field is roughly neutral. Historical precedent against repeat losers pushes down or anchors the base rate. Net: base rate 15% → adjusted to ~50-55%.
combination method:
Weighted average of sub-questions with adjustment. Sub-questions 1-3 are preconditions that are largely met or likely to be met. Sub-question 4 (formal announcement) is the binding constraint and most directly maps to resolution. Final estimate draws heavily on sq4 but incorporates the strong precondition signals.
final:
Harris is exhibiting nearly all pre-campaign behaviors typical of someone who will run: public interest signals, early state visits, declining alternative offices, maintaining political visibility. The main risks are: (1) she ultimately decides not to run despite signals (~25% chance), (2) party establishment successfully discourages her (~10%), (3) she announces after Jan 1, 2028 deadline (~5-8%). Balancing the strong upward evidence against the historically low base rate of repeat candidacies, I estimate 55% probability of formal announcement before Jan 1, 2028.
ensemble:
{'ensemble_count': 3, 'perspectives': {'base_rate': {'p_yes': 0.55, 'confidence': 0.45}, 'evidence_driven': {'p_yes': 0.55, 'confidence': 0.45}, 'contrarian': {'p_yes': 0.55, 'confidence': 0.45}}, 'spread': 0.0, 'individual_p_yes': {'base_rate': 0.55, 'evidence_driven': 0.55, 'contrarian': 0.55}}