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

MARKETS · Settled

Will above 40000 jobs be added in September 2026?

The answer: NO

The panel of 20 models averaged 70% YES; 1 of 20 were on the right side. The crowd price when the panel was sealed: 73%.

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

THE BRIEF, WRITTEN AFTER THE OUTCOME

NO; panel leaned YES (mean 70%, median 74%) while crowd was 73%, one model called NO

The outcome was NO. The panel mean was 70% and median 74%, compared with the crowd Polymarket price at 73%. Llama 4 Maverick made the best call at 23% and was the only model to call the outcome. Most models favoured YES, including two at 88%, which missed.

  • Outcome: NO.
  • Panel mean 70%, median 74%, 20 models, 1 called it.
  • Crowd: Polymarket price frozen at 73%.
  • Only Llama 4 Maverick forecast NO at 23% and was correct.
  • Highest model forecasts were 88% (Kimi K2 and Qwen3 235B) and were far off.

EVERY FORECAST, HIGHEST FIRST

  • Kimi K288%
  • Qwen3 235B88%
  • DeepSeek R187%
  • Claude Sonnet 4.585%
  • GPT-584%
  • GPT-5 Mini82%
  • Claude Opus 4.182%
  • Perplexity Sonar Pro82%
  • DeepSeek V375%
  • Grok 4.574%
  • GPT-5 + web search74%
  • Mistral Large68%
  • Gemini 2.5 Flash65%
  • Gemini 2.5 Pro63%
  • Perplexity Sonar62%
  • Claude Sonnet 4.5 + web search58%
  • Llama 4 Maverick + web search57%
  • DeepSeek V3 + web search56%
  • Gemini 2.5 Pro + web search56%
  • Llama 4 Maverick23%

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