voidly
BY · per-country backtest

Belarus forecast vs reality

Every Voidly Sentinel shutdown-risk forecast for Belarus, plotted against what actually happened. 63 (predicted, observed) pairs from the rolling 30-day evaluation window.

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

Forecasts evaluated
63
since Apr 17
Accuracy @ 0.5
38.1%
24/63 correct
Brier score
0.539
lower is better
Observed positive rate
60%
mean predicted 13%

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 17Jun 8

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

All 63 predictions (newest first)

Eval dateForecastPred ≥ 0.5?Observed?Correct?
Jun 84.2%shutdown
Jun 74.0%shutdown
Jun 63.7%shutdown
Jun 54.7%shutdown
Jun 45.3%no event
Jun 33.4%no event
Jun 23.7%shutdown
Jun 131.5%shutdown
Jun 13.8%shutdown
May 3128.4%shutdown
May 313.9%shutdown
May 305.5%shutdown
May 294.6%shutdown
May 2880.0%shutdown
May 2826.0%shutdown
May 284.5%shutdown
May 2760.0%shutdown
May 273.3%shutdown
May 263.1%no event
May 2529.8%no event
May 255.5%no event
May 2420.5%no event
May 245.2%no event
May 2360.0%no event
May 2361.4%no event
May 2280.0%no event
May 2271.8%no event
May 2218.5%no event
May 217.0%no event
May 207.2%shutdown
May 197.3%shutdown
May 186.1%shutdown
May 174.4%shutdown
May 164.7%shutdown
May 153.4%shutdown
May 145.2%no event
May 133.6%no event
May 124.5%no event
May 114.2%no event
May 103.7%no event
May 94.5%no event
May 82.7%no event
May 74.4%no event
May 63.4%no event
May 54.6%no event
May 43.8%no event
May 34.0%no event
May 24.7%shutdown
May 13.0%shutdown
Apr 304.6%shutdown
Apr 295.0%shutdown
Apr 284.3%shutdown
Apr 274.4%shutdown
Apr 263.8%shutdown
Apr 255.0%shutdown
Apr 244.8%shutdown
Apr 233.3%shutdown
Apr 224.6%shutdown
Apr 213.7%shutdown
Apr 204.9%shutdown

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