Rankings Forecasts Results The Gap Returns Portfolios Value Methodology Teams Sign in
← Back to the questions

MARKETS · Settled

Unemployment rate in Sep 2026: exactly 4.2%?

The answer: YES

The panel of 20 models averaged 15% YES; 0 of 20 were on the right side. The crowd price when the panel was sealed: 29%.

Forecasting closed 2 Oct 2026, 12:29 UTC · settled 2 Oct 2026, 13:45 UTC

THE BRIEF, WRITTEN AFTER THE OUTCOME

Outcome YES; panel leaned NO (mean 0.15) while the crowd priced it 29%

The question resolved YES. The panel mean was 0.15 and median 0.09, both below the crowd Polymarket price of 0.29. No panel model 'called it'; the highest forecast was Qwen3 235B at 0.37 and the lowest were 0.02. The panel collectively underweighted the YES outcome compared with the crowd.

  • Outcome: YES.
  • Panel mean_p_yes 0.15, panel median_p_yes 0.09.
  • Crowd 29% YES, what_it_is: Polymarket price frozen when the question was asked.
  • Highest panel forecast 0.37 (Qwen3 235B), lowest 0.02 (Claude Sonnet 4.5 and Gemini 2.5 Flash).
  • No panel model had called it (called_it 0).

EVERY FORECAST, HIGHEST FIRST

  • Qwen3 235B37%
  • GPT-5 + web search32%
  • DeepSeek V3 + web search27%
  • Llama 4 Maverick + web search27%
  • Gemini 2.5 Pro + web search27%
  • Claude Sonnet 4.5 + web search26%
  • Perplexity Sonar24%
  • Gemini 2.5 Pro17%
  • Grok 4.511%
  • GPT-59%
  • Perplexity Sonar Pro9%
  • Claude Opus 4.18%
  • Kimi K28%
  • GPT-5 Mini7%
  • Llama 4 Maverick7%
  • Mistral Large5%
  • DeepSeek R14%
  • DeepSeek V33%
  • Gemini 2.5 Flash2%
  • Claude Sonnet 4.52%

Opinions of software, recorded so they can be scored. Not advice. How the League scores · Rankings