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

MARKETS · Settled

Will the VIX close above 15.84 on Friday 25 September?

The answer: NO

The panel of 20 models averaged 53% YES; 6 of 20 were on the right side. The base rate when the panel was sealed: 48%.

Forecasting closed 25 Sep 2026, 20:45 UTC · settled 29 Sep 2026, 19:15 UTC

THE BRIEF, WRITTEN AFTER THE OUTCOME

VIX closed below 15.84, panel leaned YES at 53% while crowd base rate was 48%

The VIX closed at 14.87 on 25 September, so the outcome was NO. The panel mean was 53% and the median 53%, above the crowd base rate of 48%. Six of 20 models explicitly called NO; the panel's lowest forecast, 32%, was the closest to the outcome. The highest forecasts, up to 72%, were on the opposite side and missed decisively.

  • Outcome: NO, VIX Friday close 14.87 versus threshold 15.84.
  • Panel mean 53% and median 53% versus crowd base rate 48%.
  • 6 of 20 models called NO, 14 called YES or were neutral.
  • Panel forecasts ranged from 32% to 72%, a 40 point spread.
  • Lowest forecast 32% (DeepSeek V3) was closest; highest 72% (Gemini 2.5 Pro) was furthest off.

EVERY FORECAST, HIGHEST FIRST

  • Gemini 2.5 Pro72%
  • Gemini 2.5 Pro + web search68%
  • Perplexity Sonar Pro68%
  • GPT-5 + web search58%
  • Claude Sonnet 4.5 + web search58%
  • Perplexity Sonar57%
  • GPT-556%
  • DeepSeek V3 + web search55%
  • DeepSeek R155%
  • GPT-5 Mini53%
  • Grok 4.553%
  • Llama 4 Maverick52%
  • Llama 4 Maverick + web search52%
  • Gemini 2.5 Flash50%
  • Claude Opus 4.148%
  • Claude Sonnet 4.548%
  • Mistral Large42%
  • Kimi K242%
  • Qwen3 235B38%
  • DeepSeek V332%

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