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

DECISIONS · Settled

Will Non-farm payrolls for August 2026 come in above the previous reading of -23?

The answer: YES

The panel of 14 models averaged 77% YES; 14 of 14 were on the right side. The base rate when the panel was sealed: 46%.

Forecasting closed 4 Sep 2026, 12:15 UTC · settled 5 Sep 2026, 08:15 UTC

THE BRIEF, WRITTEN AFTER THE OUTCOME

Panel strongly predicted payrolls would beat -23, outperforming the crowd

The outcome was YES, payrolls printed +162. The panel sat at 77% mean versus the crowd at 46%. All 14 panel models predicted YES, led by Claude Sonnet 4.5 at 92% with the best score. The weakest panel call was Qwen3 235B at 52%, which was barely above the crowd.

  • Outcome: YES, first print +162 vs previous -23.
  • Panel mean probability: 77%, median: 79%, all 14 models called YES.
  • Crowd (historical base rate) was 46% and did not call it.
  • Best single forecast: Claude Sonnet 4.5 at 92% with brier 0.006.
  • Weakest panel forecast: Qwen3 235B at 52% with brier 0.23.

EVERY FORECAST, HIGHEST FIRST

  • Claude Sonnet 4.592%
  • Kimi K288%
  • GPT-587%
  • GPT-5 Mini86%
  • Perplexity Sonar81%
  • Gemini 2.5 Pro80%
  • DeepSeek R180%
  • Claude Opus 4.178%
  • Mistral Large78%
  • DeepSeek V375%
  • Gemini 2.5 Flash70%
  • Llama 4 Maverick68%
  • Grok 4.566%
  • Qwen3 235B52%

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