SY · per-country backtest
Syria 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
36.1%
52/144 correct
Brier score
0.376
lower is better
Observed positive rate
11%
mean predicted 42%
All 144 predictions (newest first)
| Eval date | Forecast | Pred ≥ 0.5? | Observed? | Correct? |
|---|---|---|---|---|
| Sep 7 | 73.6% | ↑ | no event | ✗ |
| Sep 6 | 74.4% | ↑ | no event | ✗ |
| Sep 5 | 69.4% | ↑ | no event | ✗ |
| Sep 4 | 73.2% | ↑ | no event | ✗ |
| Sep 3 | 71.0% | ↑ | no event | ✗ |
| Sep 2 | 89.0% | ↑ | no event | ✗ |
| Sep 1 | 85.9% | ↑ | no event | ✗ |
| Aug 31 | 89.6% | ↑ | no event | ✗ |
| Aug 30 | 80.3% | ↑ | no event | ✗ |
| Aug 29 | 75.8% | ↑ | no event | ✗ |
| Aug 28 | 81.0% | ↑ | no event | ✗ |
| Aug 27 | 75.9% | ↑ | no event | ✗ |
| Aug 26 | 78.2% | ↑ | no event | ✗ |
| Aug 25 | 81.6% | ↑ | no event | ✗ |
| Aug 24 | 81.0% | ↑ | no event | ✗ |
| Aug 23 | 70.0% | ↑ | no event | ✗ |
| Aug 22 | 78.3% | ↑ | no event | ✗ |
| Aug 21 | 56.5% | — | no event | ✓ |
| Aug 20 | 56.3% | — | no event | ✓ |
| Aug 19 | 70.8% | ↑ | no event | ✗ |
| Aug 18 | 72.5% | ↑ | no event | ✗ |
| Aug 17 | 63.8% | — | no event | ✓ |
| Aug 16 | 72.8% | ↑ | no event | ✗ |
| Aug 15 | 71.4% | ↑ | no event | ✗ |
| Aug 14 | 64.9% | — | no event | ✓ |
| Aug 13 | 70.3% | ↑ | no event | ✗ |
| Aug 12 | 57.6% | ↑ | no event | ✗ |
| Aug 11 | 65.5% | ↑ | no event | ✗ |
| Aug 10 | 63.2% | ↑ | no event | ✗ |
| Aug 9 | 70.5% | ↑ | no event | ✗ |
| Aug 8 | 69.1% | ↑ | no event | ✗ |
| Aug 7 | 73.0% | ↑ | no event | ✗ |
| Aug 6 | 81.5% | ↑ | no event | ✗ |
| Aug 5 | 74.3% | ↑ | no event | ✗ |
| Aug 4 | 74.2% | ↑ | no event | ✗ |
| Aug 3 | 4.1% | — | no event | ✓ |
| Aug 2 | 82.3% | ↑ | no event | ✗ |
| Aug 1 | 79.2% | ↑ | no event | ✗ |
| Jul 31 | 81.5% | ↑ | no event | ✗ |
| Jul 30 | 85.5% | ↑ | no event | ✗ |
| Jul 29 | 89.7% | ↑ | no event | ✗ |
| Jul 28 | 77.1% | ↑ | no event | ✗ |
| Jul 27 | 75.3% | ↑ | no event | ✗ |
| Jul 26 | 77.8% | ↑ | no event | ✗ |
| Jul 25 | 80.2% | ↑ | no event | ✗ |
| Jul 24 | 87.4% | ↑ | no event | ✗ |
| Jul 23 | 65.3% | ↑ | no event | ✗ |
| Jul 22 | 72.1% | ↑ | no event | ✗ |
| Jul 21 | 73.2% | ↑ | no event | ✗ |
| Jul 20 | 69.0% | ↑ | no event | ✗ |
| Jul 19 | 70.9% | ↑ | no event | ✗ |
| Jul 18 | 65.3% | ↑ | no event | ✗ |
| Jul 17 | 71.6% | ↑ | no event | ✗ |
| Jul 16 | 63.6% | ↑ | no event | ✗ |
| Jul 15 | 72.2% | ↑ | no event | ✗ |
| Jul 14 | 70.3% | ↑ | no event | ✗ |
| Jul 13 | 65.8% | ↑ | no event | ✗ |
| Jul 12 | 70.4% | ↑ | no event | ✗ |
| Jul 11 | 66.8% | ↑ | no event | ✗ |
| Jul 10 | 61.7% | ↑ | 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/sy — current 7-day calibrated forecast for Syria with SHAP drivers
- /sy — Syria 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 Syria, plotted against what actually happened. 144 (predicted, observed) pairs from the rolling 30-day evaluation window.