ET · per-country backtest
Ethiopia forecast vs reality
Every Voidly Sentinel shutdown-risk forecast for Ethiopia, 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
0.211
lower is better
Observed positive rate
17%
mean predicted 26%
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.
Y axis: forecast probability · Faint tick: distance to observed outcome (0 or 1)
All 112 predictions (newest first)
| Eval date | Forecast | Pred ≥ 0.5? | Observed? | Correct? |
|---|---|---|---|---|
| Aug 6 | 25.3% | — | no event | ✓ |
| Aug 5 | 20.4% | — | no event | ✓ |
| Aug 4 | 23.3% | — | no event | ✓ |
| Aug 3 | 22.7% | — | no event | ✓ |
| Aug 2 | 14.5% | — | no event | ✓ |
| Aug 1 | 19.6% | — | no event | ✓ |
| Jul 31 | 17.2% | — | no event | ✓ |
| Jul 30 | 14.7% | — | no event | ✓ |
| Jul 29 | 28.6% | — | no event | ✓ |
| Jul 28 | 35.6% | — | no event | ✓ |
| Jul 27 | 28.3% | — | no event | ✓ |
| Jul 26 | 35.1% | — | no event | ✓ |
| Jul 25 | 29.9% | — | no event | ✓ |
| Jul 24 | 43.8% | — | no event | ✓ |
| Jul 23 | 40.4% | — | no event | ✓ |
| Jul 22 | 40.8% | — | no event | ✓ |
| Jul 21 | 54.8% | — | no event | ✓ |
| Jul 20 | 3.8% | — | no event | ✓ |
| Jul 19 | 45.4% | — | no event | ✓ |
| Jul 18 | 69.8% | — | no event | ✓ |
| Jul 17 | 57.6% | — | no event | ✓ |
| Jul 16 | 54.1% | — | no event | ✓ |
| Jul 15 | 58.5% | — | no event | ✓ |
| Jul 14 | 62.5% | — | no event | ✓ |
| Jul 13 | 60.6% | — | no event | ✓ |
| Jul 12 | 68.2% | — | no event | ✓ |
| Jul 11 | 67.0% | ↑ | no event | ✗ |
| Jul 10 | 59.0% | ↑ | no event | ✗ |
| Jul 9 | 59.0% | ↑ | no event | ✗ |
| Jul 8 | 71.3% | ↑ | shutdown | ✓ |
| Jul 7 | 69.3% | ↑ | shutdown | ✓ |
| Jul 6 | 72.7% | ↑ | shutdown | ✓ |
| Jul 5 | 71.1% | ↑ | shutdown | ✓ |
| Jul 4 | 73.2% | ↑ | shutdown | ✓ |
| Jul 3 | 61.2% | ↑ | no event | ✗ |
| Jul 2 | 69.3% | ↑ | no event | ✗ |
| Jul 1 | 75.6% | ↑ | no event | ✗ |
| Jun 30 | 66.0% | ↑ | no event | ✗ |
| Jun 29 | 66.1% | ↑ | no event | ✗ |
| Jun 28 | 69.1% | ↑ | no event | ✗ |
| Jun 27 | 68.0% | ↑ | no event | ✗ |
| Jun 26 | 73.3% | ↑ | no event | ✗ |
| Jun 25 | 62.4% | ↑ | no event | ✗ |
| Jun 24 | 61.0% | ↑ | no event | ✗ |
| Jun 23 | 59.7% | ↑ | no event | ✗ |
| Jun 22 | 68.6% | ↑ | no event | ✗ |
| Jun 21 | 4.9% | — | no event | ✓ |
| Jun 20 | 3.4% | — | no event | ✓ |
| Jun 19 | 4.1% | — | no event | ✓ |
| Jun 18 | 3.6% | — | no event | ✓ |
| Jun 17 | 4.5% | — | no event | ✓ |
| Jun 16 | 4.4% | — | no event | ✓ |
| Jun 15 | 4.3% | — | no event | ✓ |
| Jun 14 | 5.2% | ↑ | no event | ✗ |
| Jun 13 | 3.9% | — | no event | ✓ |
| Jun 12 | 4.2% | — | no event | ✓ |
| Jun 11 | 4.2% | — | no event | ✓ |
| Jun 10 | 5.3% | ↑ | no event | ✗ |
| Jun 9 | 3.9% | — | no event | ✓ |
| Jun 8 | 5.5% | ↑ | 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
- /atlas/forecast/et — current 7-day calibrated forecast for Ethiopia with SHAP drivers
- /et — Ethiopia country page (current state, recent incidents)
- /sentinel/backtest — global reliability diagram + per-country comparison table
- Calibration refit writeup