voidly
RU · per-country backtest

Russia forecast vs reality

Every Voidly Sentinel shutdown-risk forecast for Russia, plotted against what actually happened. 112 (predicted, observed) pairs from the rolling 30-day evaluation window.

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

Forecasts evaluated
112
since Apr 17
Accuracy @ 0.5
75.9%
85/112 correct
Brier score
0.325
lower is better
Observed positive rate
58%
mean predicted 23%

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 6

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

All 112 predictions (newest first)

Eval dateForecastPred ≥ 0.5?Observed?Correct?
Aug 631.8%no event
Aug 529.8%no event
Aug 447.0%no event
Aug 325.6%no event
Aug 249.0%no event
Aug 150.4%no event
Jul 3163.4%shutdown
Jul 3061.3%shutdown
Jul 2958.7%shutdown
Jul 2865.9%shutdown
Jul 2768.9%shutdown
Jul 2676.4%shutdown
Jul 2551.2%shutdown
Jul 2449.6%shutdown
Jul 2350.7%shutdown
Jul 2258.3%shutdown
Jul 2173.0%shutdown
Jul 2051.2%shutdown
Jul 1965.4%shutdown
Jul 1850.1%shutdown
Jul 1728.9%shutdown
Jul 1635.7%shutdown
Jul 1529.5%shutdown
Jul 1450.3%shutdown
Jul 1330.1%shutdown
Jul 1231.2%shutdown
Jul 1131.6%shutdown
Jul 1034.0%shutdown
Jul 948.1%shutdown
Jul 848.1%shutdown
Jul 738.5%shutdown
Jul 639.6%shutdown
Jul 551.3%shutdown
Jul 451.3%shutdown
Jul 377.5%shutdown
Jul 251.2%shutdown
Jul 158.5%shutdown
Jun 3077.3%shutdown
Jun 2946.1%shutdown
Jun 2867.2%shutdown
Jun 2762.9%shutdown
Jun 2650.6%shutdown
Jun 2554.3%shutdown
Jun 2449.4%no event
Jun 2348.0%no event
Jun 2258.8%no event
Jun 214.5%no event
Jun 202.4%no event
Jun 194.9%no event
Jun 182.2%no event
Jun 174.4%no event
Jun 163.2%no event
Jun 153.5%no event
Jun 143.7%no event
Jun 134.8%no event
Jun 125.0%no event
Jun 114.9%no event
Jun 104.5%no event
Jun 94.7%no event
Jun 85.0%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