voidly
IN · per-country backtest

India forecast vs reality

Every Voidly Sentinel shutdown-risk forecast for India, plotted against what actually happened. 69 (predicted, observed) pairs from the rolling 30-day evaluation window.

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

Forecasts evaluated
69
since Apr 17
Accuracy @ 0.5
55.1%
38/69 correct
Brier score
0.741
lower is better
Observed positive rate
97%
mean predicted 14%

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 69 predictions (newest first)

Eval dateForecastPred ≥ 0.5?Observed?Correct?
Jun 84.4%shutdown
Jun 74.3%shutdown
Jun 65.3%shutdown
Jun 53.8%shutdown
Jun 45.4%no event
Jun 32.9%no event
Jun 22.7%shutdown
Jun 141.3%shutdown
Jun 15.1%shutdown
May 3141.9%shutdown
May 313.2%shutdown
May 3040.4%shutdown
May 302.9%shutdown
May 2944.8%shutdown
May 292.4%shutdown
May 2860.0%shutdown
May 2844.9%shutdown
May 284.8%shutdown
May 2760.0%shutdown
May 2743.7%shutdown
May 275.0%shutdown
May 2646.5%shutdown
May 264.5%shutdown
May 2539.6%shutdown
May 253.6%shutdown
May 2439.8%shutdown
May 245.6%shutdown
May 2360.0%shutdown
May 2335.4%shutdown
May 234.7%shutdown
May 2260.0%shutdown
May 2241.2%shutdown
May 225.7%shutdown
May 2134.4%shutdown
May 216.3%shutdown
May 207.5%shutdown
May 197.0%shutdown
May 187.1%shutdown
May 175.4%shutdown
May 165.2%shutdown
May 158.0%shutdown
May 146.1%shutdown
May 134.4%shutdown
May 124.6%shutdown
May 112.8%shutdown
May 104.0%shutdown
May 94.6%shutdown
May 84.3%shutdown
May 75.9%shutdown
May 65.3%shutdown
May 53.3%shutdown
May 44.3%shutdown
May 34.1%shutdown
May 21.9%shutdown
May 12.7%shutdown
Apr 304.2%shutdown
Apr 292.4%shutdown
Apr 285.4%shutdown
Apr 274.1%shutdown
Apr 264.4%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