EARLY RESULTS: 2 DAYS AHEAD

Registered 2026-07-16 · protocol 030-v1 · updated SEPTEMBER 13, 2026

Across 138 settled calls, ValuCast moved about 2 days before the public board. That is encouraging, but it is still too early to declare a win.

We timestamp each call and compare it with the public board. When the board later moves our way, we count how much earlier we were. When it moves against us, those days count against us. Calls we change and calls still waiting remain visible below — nothing disappears.

Preview graphic Download PNG

Anticipation Score

+2d current median timing versus the public board

95% CI [-3, +6] days · n = 138 · leading, not yet significant

Compared with Aggregate public consensus (1 registered cohort) each player consensus requires at least 2 public boards · it never changes our model score

76 Moved our way the public board later agreed

62 Moved against us the call went the other way

267 Calls we changed 59% of all registered calls

68 Still waiting not settled yet

454 claims registered · 138 scored into the pool · 316 excluded (clean retractions, benign expiries, still open) · 0 unscored or unclassified. Every registered claim appears in exactly one of these — the funnel partitions the whole pool.

ValuCast scores are computed without any consensus input; external source ranks are carried as context only and never feed the model. This independence is validator-enforced.