voidly
ID · per-country backtest

Indonesia forecast vs reality

Every Voidly Sentinel shutdown-risk forecast for Indonesia, 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
51.8%
58/112 correct
Brier score
0.315
lower is better
Observed positive rate
33%
mean predicted 4%

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 61.5%no event
Aug 51.5%no event
Aug 41.2%no event
Aug 31.4%no event
Aug 21.2%no event
Aug 11.0%shutdown
Jul 311.0%shutdown
Jul 300.8%shutdown
Jul 291.1%shutdown
Jul 281.2%shutdown
Jul 271.1%no event
Jul 261.1%no event
Jul 251.1%no event
Jul 241.3%no event
Jul 235.7%no event
Jul 227.3%no event
Jul 217.5%shutdown
Jul 203.9%shutdown
Jul 193.5%shutdown
Jul 182.4%shutdown
Jul 172.9%shutdown
Jul 161.5%no event
Jul 153.1%no event
Jul 141.9%shutdown
Jul 131.3%shutdown
Jul 120.9%shutdown
Jul 110.8%shutdown
Jul 100.9%shutdown
Jul 90.6%shutdown
Jul 84.4%shutdown
Jul 71.0%shutdown
Jul 60.9%no event
Jul 50.4%no event
Jul 41.2%no event
Jul 31.1%no event
Jul 21.2%no event
Jul 11.5%no event
Jun 301.2%no event
Jun 291.1%no event
Jun 281.3%no event
Jun 271.4%no event
Jun 261.2%no event
Jun 251.1%no event
Jun 240.8%no event
Jun 230.9%no event
Jun 220.7%no event
Jun 214.7%no event
Jun 203.9%no event
Jun 194.5%no event
Jun 186.0%no event
Jun 173.6%no event
Jun 166.1%no event
Jun 154.8%no event
Jun 144.4%no event
Jun 134.5%no event
Jun 125.2%no event
Jun 115.8%no event
Jun 104.9%no event
Jun 957.9%no event
Jun 84.5%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