base rate:
No major Hungarian party with 30%+ vote share has ever won fewer than 50 seats in any post-1990 election. Fidesz has historically won 133-263 seats since 2010. However, 2026 represents an unprecedented collapse for Fidesz - their worst result potentially ever. The base rate for a governing party losing this badly and still getting >50 seats depends heavily on the vote share they actually achieved.
evidence updates:
The election occurred April 12, 2026. TISZA won a two-thirds supermajority (≥133 seats). This is confirmed. Medián's pre-election projection of 49-55 Fidesz seats is the most critical data point - it puts the threshold squarely in play. STRONG evidence UP: the proportional list component provides a floor of ~28-33 seats at 30%+ vote share. STRONG evidence UP: even in a landslide loss, Fidesz would win some SMDs in strongholds. MODERATE evidence DOWN: the Medián lower bound of 49 means falling below 50 is plausible. MODERATE evidence DOWN: the SMD system amplifies the winner's advantage, potentially leaving Fidesz with very few district wins. NET: The probability centers around 78-85% for >50 seats, reflecting that while it's likely, the Medián range creates genuine uncertainty.
combination method:
Direct estimation overriding the mechanical weighted average, because sub-question 2 is not relevant to the actual resolution criterion (>50 seats). The key inputs are: (1) confirmed TISZA supermajority, (2) Medián projection of 49-55 Fidesz seats, (3) electoral system mechanics. The midpoint of Medián's range (52) is above 50, and the range itself suggests ~60-70% of outcomes fall above 50. Adding the slight historical bias toward polls underestimating the incumbent's floor, I arrive at ~82%.
final:
I estimate 82% probability that Fidesz-KDNP won above 50 seats. The election has occurred. TISZA's supermajority is confirmed. The critical unknown is whether Fidesz landed at 49, 50, 51, or somewhere in the 49-55 range projected by Medián. The midpoint and majority of that range is above 50, and structural factors (proportional seats, rural strongholds) suggest >50 is more likely than not. But this is genuinely uncertain - the threshold is right at the edge of projections. My confidence is moderate (0.55) because I lack the actual certified seat count.
ensemble:
{'ensemble_count': 3, 'perspectives': {'base_rate': {'p_yes': 0.82, 'confidence': 0.55}, 'evidence_driven': {'p_yes': 0.87, 'confidence': 0.45}, 'contrarian': {'p_yes': 0.85, 'confidence': 0.55}}, 'spread': 0.05, 'individual_p_yes': {'base_rate': 0.82, 'evidence_driven': 0.87, 'contrarian': 0.85}}