SD · per-country backtest
Sudan forecast vs reality
Compare published forecasts with observed outcomes for this country.
Updated every 30 min · CC BY 4.0 · Raw JSON · Current forecast →
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)
Forecasts evaluated
144
since Apr 17
Accuracy @ 0.5
65.3%
94/144 correct
Brier score
0.136
lower is better
Observed positive rate
5%
mean predicted 26%
All 144 predictions (newest first)
| Eval date | Forecast | Pred ≥ 0.5? | Observed? | Correct? |
|---|---|---|---|---|
| Sep 7 | 1.2% | — | no event | ✓ |
| Sep 6 | 1.4% | — | no event | ✓ |
| Sep 5 | 1.3% | — | no event | ✓ |
| Sep 4 | 1.3% | — | no event | ✓ |
| Sep 3 | 1.3% | — | no event | ✓ |
| Sep 2 | 1.3% | — | no event | ✓ |
| Sep 1 | 1.3% | — | no event | ✓ |
| Aug 31 | 1.3% | — | no event | ✓ |
| Aug 30 | 2.5% | — | no event | ✓ |
| Aug 29 | 2.5% | — | no event | ✓ |
| Aug 28 | 52.3% | ↑ | no event | ✗ |
| Aug 27 | 52.3% | ↑ | no event | ✗ |
| Aug 26 | 52.3% | ↑ | no event | ✗ |
| Aug 25 | 52.3% | ↑ | no event | ✗ |
| Aug 24 | 52.3% | ↑ | no event | ✗ |
| Aug 23 | 62.6% | ↑ | no event | ✗ |
| Aug 22 | 62.6% | ↑ | no event | ✗ |
| Aug 21 | 60.1% | ↑ | no event | ✗ |
| Aug 20 | 60.1% | ↑ | no event | ✗ |
| Aug 19 | 60.1% | ↑ | no event | ✗ |
| Aug 18 | 15.6% | — | no event | ✓ |
| Aug 17 | 15.6% | — | no event | ✓ |
| Aug 16 | 8.2% | — | no event | ✓ |
| Aug 15 | 3.9% | — | no event | ✓ |
| Aug 14 | 3.9% | — | no event | ✓ |
| Aug 13 | 3.9% | — | no event | ✓ |
| Aug 12 | 48.2% | ↑ | no event | ✗ |
| Aug 11 | 48.2% | ↑ | no event | ✗ |
| Aug 10 | 48.2% | ↑ | no event | ✗ |
| Aug 9 | 53.2% | ↑ | no event | ✗ |
| Aug 8 | 76.1% | ↑ | no event | ✗ |
| Aug 7 | 76.1% | ↑ | no event | ✗ |
| Aug 6 | 76.1% | ↑ | no event | ✗ |
| Aug 5 | 71.5% | ↑ | no event | ✗ |
| Aug 4 | 71.5% | ↑ | no event | ✗ |
| Aug 3 | 71.5% | ↑ | no event | ✗ |
| Aug 2 | 73.3% | ↑ | no event | ✗ |
| Aug 1 | 73.3% | — | no event | ✓ |
| Jul 31 | 73.3% | ↑ | no event | ✗ |
| Jul 30 | 71.6% | ↑ | no event | ✗ |
| Jul 29 | 66.5% | — | no event | ✓ |
| Jul 28 | 17.9% | — | no event | ✓ |
| Jul 27 | 17.9% | — | no event | ✓ |
| Jul 26 | 19.3% | — | no event | ✓ |
| Jul 25 | 16.2% | — | no event | ✓ |
| Jul 24 | 16.2% | — | no event | ✓ |
| Jul 23 | 16.2% | — | no event | ✓ |
| Jul 22 | 12.7% | — | no event | ✓ |
| Jul 21 | 12.7% | — | no event | ✓ |
| Jul 20 | 12.7% | — | no event | ✓ |
| Jul 19 | 14.2% | — | no event | ✓ |
| Jul 18 | 14.2% | — | no event | ✓ |
| Jul 17 | 14.2% | — | no event | ✓ |
| Jul 16 | 21.2% | — | no event | ✓ |
| Jul 15 | 21.2% | — | no event | ✓ |
| Jul 14 | 17.0% | — | no event | ✓ |
| Jul 13 | 19.2% | — | no event | ✓ |
| Jul 12 | 7.2% | — | no event | ✓ |
| Jul 11 | 7.2% | — | no event | ✓ |
| Jul 10 | 6.7% | — | 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/sd — current 7-day calibrated forecast for Sudan with SHAP drivers
- /sd — Sudan country page (current state, recent incidents)
- /sentinel/backtest — global reliability diagram + per-country comparison table
- Calibration refit writeup
About this view
Every Voidly Sentinel shutdown-risk forecast for Sudan, plotted against what actually happened. 144 (predicted, observed) pairs from the rolling 30-day evaluation window.