649
Forecasts published
All time
386
Resolved forecasts
59% of total
37%
Mean accuracy
Across resolved
176
Confirmed
46% of resolved
95
Partially confirmed
25% of resolved
115
Missed
30% of resolved
0
Forecasts edited
Always published as-is
We're projecting rising demand for Python and cooling demand for Platform Engineer — 386 past forecasts scored so far at 37% mean accuracy.
Forecasts are immutable once published — never edited, never quietly withdrawn. When the target period arrives each one is scored against the observed count, and the misses stay on this page alongside the hits.
Most important active forecasts, by evidence — target week Nov 9–15, 2026
Down from 75 postings the week of Oct 12–18, 2026.
~15postings
Expected Nov 9–15, 2026
Down from 41 postings the week of Oct 12–18, 2026.
~14postings
Expected Nov 9–15, 2026
Down from 61 postings the week of Oct 12–18, 2026.
~15postings
Expected Nov 9–15, 2026
Down from 93 postings the week of Oct 12–18, 2026.
~14postings
Expected Nov 9–15, 2026
Forecast accuracy by resolution week
30%
Mean accuracy · 386 resolved
Outcomes across 386 resolved forecasts
We keep our misses on the table. Forecast quality is judged on all 386, not just the wins.
Every forecast follows the same six steps
176 forecasts · sorted by target week
Every forecast is published. Every forecast is locked. Every forecast is eventually scored. No editing. No disappearing misses. Strategic Honesty.