voidly
UZ · per-country backtest

Uzbekistan forecast vs reality

Every Voidly Sentinel shutdown-risk forecast for Uzbekistan, plotted against what actually happened. 118 (predicted, observed) pairs from the rolling 30-day evaluation window.

Updated every 30 min · CC BY 4.0 · Raw JSON · Current forecast →

Forecasts evaluated
118
since Apr 17
Accuracy @ 0.5
36.4%
43/118 correct
Brier score
0.441
lower is better
Observed positive rate
19%
mean predicted 53%

Forecast time series

Blue line: forecast probability. Green ✓ markers: forecast was right. Red ✗ markers: forecast was wrong. Dashed line at 0.5 is the binary decision threshold.

0.000.250.500.751.00threshold 0.50Apr 17Aug 12

Y axis: forecast probability · Faint tick: distance to observed outcome (0 or 1)

All 118 predictions (newest first)

Eval dateForecastPred ≥ 0.5?Observed?Correct?
Aug 1254.3%no event
Aug 1159.2%no event
Aug 1066.1%no event
Aug 963.8%no event
Aug 843.1%no event
Aug 755.8%no event
Aug 641.7%no event
Aug 538.5%no event
Aug 457.5%no event
Aug 371.9%no event
Aug 268.3%no event
Aug 180.8%no event
Jul 3177.2%no event
Jul 3063.2%no event
Jul 2963.9%shutdown
Jul 2860.8%shutdown
Jul 2762.0%shutdown
Jul 2662.8%shutdown
Jul 2568.1%no event
Jul 2465.7%no event
Jul 2364.5%no event
Jul 2272.5%no event
Jul 2174.9%no event
Jul 2087.6%no event
Jul 1966.8%no event
Jul 1865.8%no event
Jul 1764.9%no event
Jul 1671.0%no event
Jul 1569.1%no event
Jul 1472.6%no event
Jul 1376.3%no event
Jul 1269.6%no event
Jul 1175.9%no event
Jul 1073.0%no event
Jul 969.5%no event
Jul 868.9%no event
Jul 777.4%no event
Jul 675.9%no event
Jul 569.7%no event
Jul 473.8%no event
Jul 375.9%no event
Jul 273.8%no event
Jul 182.9%no event
Jun 3082.0%no event
Jun 2979.2%no event
Jun 2878.3%no event
Jun 2783.3%no event
Jun 2683.1%no event
Jun 2586.9%no event
Jun 2485.9%no event
Jun 2380.5%no event
Jun 2277.6%no event
Jun 2163.5%no event
Jun 2063.9%no event
Jun 1962.8%no event
Jun 1865.2%no event
Jun 1763.9%no event
Jun 1664.8%no event
Jun 1563.1%no event
Jun 1464.7%no event

How to read this

  • Each row is one historical forecast. We made the prediction at eval_date, then waited the 7-day horizon, then graded against the observed outcome.
  • Forecast % is the calibrated probability we published that day. Post-recalibration (2026-05-20) these now match actual observed rates much better — see the refit finding.
  • Pred ≥ 0.5 shows the binary alert decision. Note we usually fire alerts at a lower threshold (see /v1/sentinel/global_heatmap for the live cutoff) — the 0.5 column here is for backtest scoring consistency with the global confusion matrix.
  • Correct = both (pred ≥ 0.5) and (observed) agree.

Related