voidly
PH · per-country backtest

Philippines forecast vs reality

Every Voidly Sentinel shutdown-risk forecast for Philippines, 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
58.9%
66/112 correct
Brier score
0.028
lower is better
Observed positive rate
0%
mean predicted 11%

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 63.7%no event
Aug 54.6%no event
Aug 44.9%no event
Aug 34.4%no event
Aug 24.3%no event
Aug 14.1%no event
Jul 3136.2%no event
Jul 3035.9%no event
Jul 2933.1%no event
Jul 2833.1%no event
Jul 2730.3%no event
Jul 2634.5%no event
Jul 2536.4%no event
Jul 2435.1%no event
Jul 2334.9%no event
Jul 2235.1%no event
Jul 2133.4%no event
Jul 2030.7%no event
Jul 1936.0%no event
Jul 1835.0%no event
Jul 1734.8%no event
Jul 1636.1%no event
Jul 1536.1%no event
Jul 1432.8%no event
Jul 1333.5%no event
Jul 1236.2%no event
Jul 1134.5%no event
Jul 105.1%no event
Jul 95.6%no event
Jul 84.9%no event
Jul 74.9%no event
Jul 64.3%no event
Jul 55.3%no event
Jul 45.1%no event
Jul 35.2%no event
Jul 24.2%no event
Jul 14.2%no event
Jun 304.2%no event
Jun 293.5%no event
Jun 284.4%no event
Jun 275.6%no event
Jun 265.4%no event
Jun 255.5%no event
Jun 246.1%no event
Jun 235.2%no event
Jun 224.0%no event
Jun 215.9%no event
Jun 204.2%no event
Jun 193.7%no event
Jun 183.1%no event
Jun 175.5%no event
Jun 165.9%no event
Jun 155.1%no event
Jun 144.9%no event
Jun 133.9%no event
Jun 125.0%no event
Jun 114.8%no event
Jun 105.5%no event
Jun 93.9%no event
Jun 86.1%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