voidly
ID · per-country backtest

Indonesia forecast vs reality

Every Voidly Sentinel shutdown-risk forecast for Indonesia, plotted against what actually happened. 28 (predicted, observed) pairs from the rolling 30-day evaluation window.

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

Forecasts evaluated
28
since Apr 17
Accuracy @ 0.5
32.1%
9/28 correct
Brier score
0.627
lower is better
Observed positive rate
68%
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 17May 14

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

All 28 predictions (newest first)

Eval dateForecastPred ≥ 0.5?Observed?Correct?
May 144.9%no event
May 133.4%no event
May 125.0%no event
May 114.8%no event
May 104.6%no event
May 94.9%no event
May 83.9%no event
May 72.3%no event
May 64.0%no event
May 54.8%shutdown
May 43.4%shutdown
May 33.1%shutdown
May 24.2%shutdown
May 14.0%shutdown
Apr 304.3%shutdown
Apr 294.9%shutdown
Apr 283.8%shutdown
Apr 273.2%shutdown
Apr 265.4%shutdown
Apr 252.7%shutdown
Apr 244.6%shutdown
Apr 234.7%shutdown
Apr 223.0%shutdown
Apr 214.4%shutdown
Apr 204.6%shutdown
Apr 192.6%shutdown
Apr 183.5%shutdown
Apr 174.0%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