voidly
VN · per-country backtest

Vietnam forecast vs reality

Every Voidly Sentinel shutdown-risk forecast for Vietnam, plotted against what actually happened. 73 (predicted, observed) pairs from the rolling 30-day evaluation window.

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

Forecasts evaluated
73
since Apr 17
Accuracy @ 0.5
63.0%
46/73 correct
Brier score
0.338
lower is better
Observed positive rate
36%
mean predicted 7%

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 28

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

All 73 predictions (newest first)

Eval dateForecastPred ≥ 0.5?Observed?Correct?
Jun 2812.5%shutdown
Jun 275.7%shutdown
Jun 267.1%shutdown
Jun 254.0%no event
Jun 244.7%no event
Jun 234.7%no event
Jun 226.4%no event
Jun 215.0%no event
Jun 204.8%no event
Jun 194.8%no event
Jun 184.1%no event
Jun 172.7%no event
Jun 163.9%no event
Jun 152.3%no event
Jun 144.1%no event
Jun 133.2%no event
Jun 123.5%no event
Jun 115.2%no event
Jun 105.4%no event
Jun 92.4%no event
Jun 85.1%no event
Jun 75.4%no event
Jun 63.9%shutdown
Jun 55.7%shutdown
Jun 45.1%shutdown
Jun 35.0%shutdown
Jun 24.2%shutdown
Jun 14.9%shutdown
May 315.3%shutdown
May 303.3%no event
May 294.3%no event
May 284.0%no event
May 274.2%no event
May 264.2%no event
May 254.1%no event
May 2459.6%no event
May 2361.1%no event
May 225.9%no event
May 2168.2%no event
May 207.8%no event
May 197.6%no event
May 187.0%no event
May 174.3%no event
May 164.9%no event
May 154.3%no event
May 144.1%no event
May 133.4%no event
May 123.7%no event
May 113.3%no event
May 103.8%no event
May 94.0%no event
May 82.6%no event
May 74.5%no event
May 62.7%shutdown
May 54.0%shutdown
May 43.3%shutdown
May 35.0%shutdown
May 24.7%shutdown
May 14.8%shutdown
Apr 305.5%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